Show all abstracts
View Session
- Front Matter: Volume 7820
- Image Acquisition and Processing, Pattern Recognition, and Computer Vision in Industrial Engineering
- GIS, GPS, RS, & Wireless and Optical Communications in Industrial Engineering
- Industrial Decision Support and Simulation Systems, Sensor Technology, and Intelligent Monitoring and Control/ICT Applications
Front Matter: Volume 7820
Front Matter: Volume 7820
Show abstract
This PDF file contains the front matter associated with SPIE Proceedings Volume 7820, including the Title Page, Copyright information, Table of Contents, Introduction and the Conference Committee listing.
Image Acquisition and Processing, Pattern Recognition, and Computer Vision in Industrial Engineering
Real-time tracking system combining mixture of Gaussian model and CamShift algorithm
Show abstract
In this paper, a system combining mixture of Gaussian background model and CamShift algorithm has been designed to
detect and track automatically the moving target. For detecting the moving object, the region which the movement object
belongs to is firstly identified and extracted by mixture model of Gaussian, and the centroid of this region is determined
as the center of initializing window for tracking. The color feature of object is then extracted in the region, and the
CamShift algorithm is used to calculate t he exact location of the target and adjust the size of search window. During the
tracking, the information about the objective location is transmitted by serial communication to control the PTZ in order
to track the object to ensure it being inside of scene all along. The experiment is performed with high speed spherical
camera E588/G3-HP and validates the validity of the system.
Application of rough set for medical images data mining
Shuyan Wang,
Chunmei Wang,
Yan Chen
Show abstract
To study the application of Rough set algorithm for diagnosis breast cancer, attribute reduction strategies of rough set are
applied to the data mining of the mammography classification, proposes a medical images classifier based on association
rules. Attribute reduction strategies of rough set for medical image data mining are realized. The experiment results are
given. The experimental results show that the system performs well in accuracy, verified the great potential of rough set
in assistant medical treatment.
Implementation of 2-D DCT based on FPGA
Bao-Zeng Guo,
Li Niu,
Zhi-Ming Liu
Show abstract
Discrete Cosine Transform (DCT) plays an important role in the image and video compression, and it has been widely
used in JPEG, MPEG, H.26x. DCT being implemented by hardware is crucial to improve the speed of image
compression. This paper presents a method that 2-D DCT is implemented by FPGA, which is based on the algorithm of
row-column decomposition, and the parallel structure is used to achieve high throughput. The design is achieved by
top-down design methodology and described with Verilog HDL in RTL level. The hardware of 2-D DCT is implemented
by the FPGA EP2C35F672C8 made by ALTERA. The experiment results show that the delay time is as low as 15 ns, and
the clock frequency as high as 138.35 MHz, which can satisfy the requirements of the real-time video image
compression.
A position correction equipment of work-piece in assembly line
Show abstract
This paper presents a work-piece position correction equipment. In order to correct work-piece position, for using video
signal, this paper illustrates the method of similarity rate of labelled images to carry on object location and to calculate
work-piece's deviate angle. This method is proved to be useful and efficient in testing. This paper is new and original in
three facts. First, using similarity rate of labelled image is more convenient for this purpose than the complex
combination of sensors and machinery. Second, the algorithm of this method is easier to realize than other methods.
Third, this equipment can fit a multi-purpose situation through flexible adjusting matching labelled image. This method
in the paper can be widely used in other similar application.
Traditional Chinese painters and calligraphers' seal surface analysis research based on the evenness
Show abstract
This paper proposed an effective method to analyze Traditional Chinese Painters and Calligraphers' Seal (TCPCS)
Surface based on uniformity which is one of the art of composition descriptor. This art of composition can be applied to
recognize common two and four words seals surface layout. Firstly, analysis of several common TCPCS surface layouts,
and send it as the layout of templates; searched and marked a TCPCS boundary namely image of art composition, then
design a template matching method to achieve TCPCS surface analysis results by computing the uniformity distance
difference between the common TCPCS surface template and image of art composition. Experiment results show the
effectiveness of the method.
A new license plate extraction framework based on fast mean shift
Show abstract
License plate extraction is considered to be the most crucial step of Automatic license plate recognition (ALPR) system.
In this paper, a region-based license plate hybrid detection method is proposed to solve practical problems under
complex background in which existing large quantity of disturbing information. In this method, coarse license plate
location is carried out firstly to get the head part of a vehicle. Then a new Fast Mean Shift method based on random
sampling of Kernel Density Estimate (KDE) is adopted to segment the color vehicle images, in order to get candidate
license plate regions. The remarkable speed-up it brings makes Mean Shift segmentation more suitable for this
application. Feature extraction and classification is used to accurately separate license plate from other candidate regions.
At last, tilted license plate regulation is used for future recognition steps.
Web log mining based on improved FCM clustering algorithm
Zhijun Wang,
Runjing Zhou
Show abstract
In view of more iterative times, longer convergence time and lower accuracy of Fuzzy C-Means (FCM) clustering
algorithm, an improved FCM (IFCM) clustering algorithm is put forward. After data preprocessing of web log, fuzzy
clustering to the web log data is adopted with the help of IFCM clustering algorithm. It can provide foundation and help
for the subsequent web page optimization and personalization services.
Change mechanism and change type analysis for SAR images
Show abstract
Synthetic aperture radar (SAR) image change detection is a very complex problem like SAR imaging mechanism. Every
change detection method has its advantage and disadvantage and there are no optimal change detection approaches. If
general methods directly perform SAR image change detection, it can not obtain the satisfied results. Sometimes, for a
pair of multi-temporal SAR images, different detection methods can obtain different detected results; what's more, the
results are contrary. In order to improve SAR image change detection precision, this paper studies SAR image change
mechanism in detail and represents the change types of SAR images. The new integrative change detection algorithm is
proposed, and the real SAR image data tests the methods.
A review of salient region extraction
Hong Bao,
Difei Wang
Show abstract
Visual attention derives from biology, by means of extracting the most informative region, can also be used to enhance
the efficiency in computer vision, Reviewing studies on it, analyzing and comparing features of methods usually used in
the process of salient region extraction.
Study on the classification algorithm of degree of arteriosclerosis based on fuzzy pattern recognition
Show abstract
Pulse wave of human body contains large amount of physiological and pathological information, so the degree of
arteriosclerosis classification algorithm is study based on fuzzy pattern recognition in this paper. Taking the human's
pulse wave as the research object, we can extract the characteristic of time and frequency domain of pulse signal, and
select the parameters with a better clustering effect for arteriosclerosis identification. Moreover, the validity of
characteristic parameters is verified by fuzzy ISODATA clustering method (FISOCM). Finally, fuzzy pattern recognition
system can quantitatively distinguish the degree of arteriosclerosis with patients. By testing the 50 samples in the built
pulse database, the experimental result shows that the algorithm is practical and achieves a good classification
recognition result.
On-line print-defect detecting in an incremental subspace learning framework
Show abstract
Real-time detecting print-defect system is significant for automatically quality control in printing industry. The state-of-the-art
detecting algorithms are based on conventional template matching process and usually suffer from false alarm caused
by acceptable variations. This paper proposes a novel on-line print-defect detecting approach which uses incremental
principal component analysis to model a variety pattern with respect to the detected image itself. The algorithm is
constructed and deployed to a real-time detecting print-defect system, and the test results show that the system reduces
false alarm dramatically.
Robust watermarking in wavelet domain based on chaotic scrambling
Yanling Wang
Show abstract
Recently the protection of digital information has received significant attention within the digital media community
and a number of techniques that try to address the problem by hiding appropriate information within digital media have
been proposed. In this paper, a new watermarking method in discrete wavelet transform (DWT) domain is proposed.
Firstly, we scrambled the watermark image based chaotic sequence and mapping to avoid the block effect. Secondly, we
insert the scrambled watermark in LH2 and HL2 sub-band of the DWT of the host image to provide a good tradeoff
between the transparency and the robustness of watermarks. The experimental results show that the embedded
watermark in this way is sufficiently transparent and robust against noise and commonly used image processing methods
such as Gaussian, JPEG compression, Median filtering etc.
Animated models coarsening with local area distortion and deformation degree control
Shixue Zhang,
Yuanhao Wu,
Jinyu Zhao
Show abstract
In computer graphics applications, mesh coarsening is an important technique to alleviate the workload of visualization
processing. Compared to the extensive works on static model approximation, very little attentions have been paid to
animated models. In this paper, we propose a new method to approximate animated models with local area distortion and
deformation degree control. Our method uses an improved quadric error metric guided by a local area distortion
measurement as a basic hierarchy. Also, we define a deformation degree parameter to be embedded into the aggregated
quadric errors, so areas with large deformation during the animation can be successfully preserved. Finally, a mesh
optimization process is proposed to further reduce the geometric distortion for each frame. Our approach is fast, easy to
implement, and as a result good quality dynamic approximations with well-preserved sharp features can be generated at
any given frame.
Fabric defect detection based on textured characteristics using wavelet transform
Show abstract
In texture defect detection, the defects can be discriminated according to the distribution ranges of wavelet coefficients
between the normal and defective parts of texture images. In traditional texture defect detection methods, the normal
parts of texture images have to be trained in advance. In this paper, we propose a novel method to automatically
determine the training regions based on the characteristics exhibited by normal and defective texture images. In this way,
the detection error can be reduced because of the avoiding of environmental changes.
Edge detection based on multi-scale wavelet
Tingwan Wu,
Yihui Duan,
Baoliang Liu
Show abstract
In this paper, we present a novel wavelet-based algorithm for multi-scale edge detection. Firstly, calculate wavelet
transform coefficients of the image according to the direction of the gradient. Then scan the neighborhood of the
corresponding wavelet transform coefficients separately at three-scales, in order to position edges at a small scale and
suppress noise at a large scale. The simulation results show that the new algorithm is feasible and effective, more details
can be detected. Image clarity handled with new wavelet edge detection algorithm is superior to that of the best currently
with Canny operator. For different input image, the novel algorithm can make us get a better edge map.
A fast digital image correlation algorithm based on Hartley transform and global sum-table scheme
Show abstract
Digital image correlation method (DICM) as a flexible tool for deformation measurements has found widespread use in a
variety of fields. One major challenge in practical applications, however, lies in that this technique founded on zeronormalized
cross-correlation (ZNCC) criterion is extremely time-consuming in correlation calculation. This paper
introduces the Hartley transform and a so-called global sum-table strategy to evaluate the cross-correlation term and all
double sums in the ZNCC expression, which can dramatically reduce computational complexity of correlation searching
and therefore lead to significant computational savings. Both simulation tests and an actual example of displacement
acquisition are employed to validate the feasibility and effectiveness of the fast algorithm, which indicates that it can
improve the efficiency of the DICM calculation by ~4-14 times in comparison with the conventional algorithm. This is
crucial for realizing a high-efficiency DICM to satisfy speed requirements for time-critical applications and large-scale
data processing.
A new robust region-based ICA-SIFT shape descriptor for object recognition
Show abstract
The paper proposed a new region-based ICA-SIFT shape descriptor. It combines two methods to realize the optimal
performance: ICA to process global information and SIFT to get local features, therefore, it can describe various kinds of
shapes accurately and concisely. The ICA-SIFT shape descriptor is proved to be invariant to skewing, scaling, translation
and rotation. The main process of the ICA-SIFTSD is first to extract the canonical form of an original shape by ICA and
has eliminated any effects of skewing and affine transformation. Next, we carried out SIFT feature extraction on
canonical forms and got the ICA-SIFT shape descriptor, which is an improvement of the ICAZMSD method in [3]. We
have applied FastICA and SURF( the speedup of SIFT) that can accelerate the calculation speed of the proposed method
so that it can meet requirements of real-time applications. In the paper, we carried out a large number of experiments on
the MPEG-7-CE database, using the ICA-SIFTSD as an effective descriptor for object recognition. The experimental
results show recognition rates are 91.7% and 93.8% of simple and complex shape images respectively.
The mathematical morphology of paper's defect detection method based on multi-scale and multi-structure elements morphology
Jie Kang,
Gang Yang
Show abstract
Since mathematic morphology of paper's defect detection method based on single-structure element has the shortcoming
of obtaining incomplete and discontinued defect information, a new method of mathematical morphology defect
detection on paper image using multi-scale and multi-structure elements morphology was proposed. Firstly, on the same
scale, the multi-structure elements were adopted in edge detection on paper image, and then the edge image on one scale
was gained through synthesizing. Secondly, the edge images on the different scales were weighted summed, binarized
and suppressed the noise, and at last, the edge image of paper's defect detection was obtained using multi-scale and
multi-structure elements morphology. The results of simulation show that the algorithm in this paper has better antiinterference
and positioning accuracy, and the obtained edge is more complete.
A novel automatic segmentation of the left ventricle cavity and myocardium in MSCT data
Xingjia Wang,
Lina Dong,
Yufeng Huang,
et al.
Show abstract
The manual segmentation of 3D high resolution cardiac multi slice CT (MSCT) datasets is both labor intensive and time
consuming. Therefore, it is necessary to provide a powerful automatic/semi-automatic method to segmentation the
cardiac myocardium and cavities. In this paper a novel approach for the automatic 3D segmentation has been developed
to extract the epicardium and endocardium boundaries of the left ventricle (LV) of the heart. The segmentation of the
MSCT data is divided into three parts. The first part, which is based on nonlinear intensity transformation and bilateral
filter, paints background and smoothes slice for all real CT images; The second part, applies a cavity template mask to
extract the LV cavity coarse region from all slices using the threshold and morphologic operations; The last part
performs improved coupled level set algorithm incorporating coarse cavity contours and priors for the final
segmentation. Experimental results and 3D surface reconstruction show the efficacy and advantage of our method for the
segmentation of the left ventricle from real CT data.
Brain white matter tractography based on Riemannian manifold
Lu Meng,
Bin Liu,
Hong Zhao,
et al.
Show abstract
Diffusion tensor imaging (DTI) is the only noninvasive technique of analyzing and qualifying water molecule's
diffusion anisotropy in brain tissues. This paper presented a novel algorithm to analyze DTI and brain white matter
tractography based on Riemannian manifold. Firstly, a 3×3 symmetric positive definite covariant tensor was constructed
for each voxel using DTI, so brain white matte can be represented as a tensor field. Secondly, the tensor field was
regarded as Riemannian manifold, and the fluid motion in the tensor field was represented by Navier-Stoke equation, so
the problem of brain white matter tractography between two voxels can be transformed into the computation of smallest
distance between two points in Riemannian manifold. Finally, distances between two points in Riemannian manifold can
be represented by geodesic, and the numerical solution was based on Level-Set method, which was the brain white
matter tractography. In experiment, this paper compared our method and the traditional algorithm based on a digital DTI
phantom. The experiment result showed that our method could accurately retrieve the DTI tractography, and was more
robust than traditional algorithm.
Efficient data association for robot 3D vision-SLAM
Xiao-hua Wang,
Dai-xian Zhu
Show abstract
A new approach to vision-based simultaneous localization and mapping (SLAM) is proposed. the scale invariant
feature transform (SIFT) features is landmarks, The minimal connected dominating set(CDS) approach is used in data
association which solve the problem that the scale of data association increase with the map grows in process of SLAM.
SLAM is completed by fusing the information of binocular vision and robot pose with Extended Kalman Filter
(EKF). The system has been implemented and tested on data gathered with a mobile robot in a typical office environment.
Experiments presented demonstrate that proposed method improves the data association and in this way leads to more
accurate maps.
Wood adhesion cell segmentation scheme based on GVF-Snake model
Show abstract
In order to extract the characteristic parameters of the wood cells accurately, this paper presents an efficient scheme
for wood cell segmentation. This scheme is mainly based on GVF-Snake model and the method of image thinning.
Firstly, computing the Category Roundness of every connectivity domain is done in order to get the degree of adhesion.
Secondly, image thinning helps to get the skeleton of the cell. Finally, according to the location coordinates of skeleton
and contour, it can determine the location of segmentation. Experimental results demonstrate the scheme for precise
extraction with limited human intervention; it can also determine the correct edge of segmentation. Comparatively
speaking, the inaccuracy is rather limited.
Energy analysis of video signal and interference removal
Lin Mo,
Xun Liu,
Fu-yuan Zhang
Show abstract
This paper proposed a method of analysis the energy of video signal and an interference removal algorithm which
bases on it. First, we give the definition of signal energy, which follows by a discussion of the signal average energy and
instant energy. Next we construct self-adaptive threshold of the average energy, and then propose an interference removal
algorithm. Finally, the experiment results of single pixel show that the algorithm is very effective.
Research on location and crawling of work-piece based on binocular vision
Jian Xu,
Dai-xian Zhu
Show abstract
In order to solve the work-piece recognition, location and crawling problem in the process of manufacturing, a new
method of work-piece recognition and location based on binocular stereo vision is proposed. The SIFT algorithm is used
to extract feature of each work-piece, the template matching method is used to identify Work-piece ,and work-piece
three-dimensional information is obtained by using binocular vision, Thus robot crawling task can be implemented.
MATLAB and VC++ are used to program system software. Experimental results demonstrate the good results of this
method.
Human location and recognition for intelligent air conditioners
Bing Sun,
Ke Li,
Fei Weng,
et al.
Show abstract
Through analyzing the low resolution video captured by a single camera fixed on the air condition, this paper proposes
an approach that can automatically estimate the person's location and recognize the person's identification in real time.
Human location can be obtained by smart geometry calculation with the knowledge of the camera intrinsic parameters
and living experience. Human recognition has been found to be very difficult in reality, especially when the person is
walking at a distance in the complexity indoor conditions. For optimal performance, we use the shape feature gait energy
image (GEI) as the basis, since it isn't sensitive the noise. Then we extract more efficient features using the histograms
of oriented gradients (HOG) and do the dimensionality reduction by the coupled subspaces analysis and discriminant
analysis with tensor representation (CSA+DATER), Finally the classical Bayesian Theory is used for fusion of the result
of HOG and the result of CSA+DATER. The proposed approach is tested on our lab database to evaluate the
performance of the human location and recognition. To verify the robust of our human recognition approach especially,
CMU MoBo gait database is used. Experimental results show that the proposed approach has a high accuracy rate in both
human identification recognition and location estimation.
Spatial object classification and recognition based on symmetric fuzzy relative entropy
Yufeng Shi,
Jian Chen
Show abstract
Based on fuzzy set theory, entropy, relative entropy and fuzzy entropy, the symmetric fuzzy relative entropy(SFRE) is
presented, which not only has a full physical meaning, but also has succinct practicability. The symmetric fuzzy relative
entropy can be used to measure the divergence between different fuzzy patterns. The example demonstrates that the
symmetric fuzzy relative entropy is valid and reliable for spatial object recognition and classification, and its
classification precision is very high.
Capture of meaningful faces from motion images
Ke Li,
Bing Sun,
Yuncai Liu
Show abstract
There is an increasing need and demand for automatic, efficient and reliable key information extraction from the video in
industrial engineer. The face is one kind of the key information and usually the image is not clear in low-cost real-time
surveillance system. The definition of the image mostly depends on the energy of the high frequency. The Fast Fourier
Transform (FFT) and the Discrete Wavelet Transform (DWT) are two excellent transform tools. This paper proposes a
method using FFT and DWT to analyze the face definition in motion images. The proposed method consists of three
steps. Firstly, we preprocess the image by Histogram equalization and morphology. Secondly, we processes face
detection and tracking. Finally, we capture the clearest face. We compare FFT with DWT, and experimental results show
that DWT base on Bior3.1 is the most efficient.
Efficient approach for binocular vision-SLAM
Dai-xian Zhu
Show abstract
This paper presents an approach to binocular vision simultaneous localization and mapping (SLAM). SIFT (Scale
Invariant Feature Transform) algorithm is used to extract the Natural landmarks. But SIFT algorithm is complicated and
computation time is long. Firstly, the linear combination of cityblock distance and chessboard distance is comparability
measurement; secondly, partial features are used to matching. SLAM is completed by fusing the information of SIFT
features and robot information with EKF. Mahalanobisis distance is used in data association which solve the problem
that the scale of data association increase with the map grows in process of SLAM .The simulation experiment indicate
that the proposed method reduce computational complexity, and with high localization precision in indoor environments.
Multiple mobile robots real-time visual search algorithm
Show abstract
A multiple mobile robots visual real-time locating system is introduced, in which the global search algorithm and track
search algorithm are combined together to identify the real-time position and orientation(pose) of multiple mobile robots.
The switching strategy between the two algorithms is given to ensure the accuracy and improve retrieval speed. The grid
search approach is used to identify target while searching globally. By checking the location in the previous frame, the
maximum speed and the frame time interval, thus the track search can determine the area target robot may appear in the
next frame. Then, a new search will be performed in the certain area. The global search is used if target robot is not
found in the previous search otherwise track search will be used. With the experiment on the static and dynamic
recognition of three robots, the search method here is proved to be high precise, fast, stable and easy to extend, all the
design requirements can be well met.
An adaptive enhancement means adapt to CR medicine image
Ming-Hui Zhang,
Yao-Yu Zhang
Show abstract
It puts forward CR medicine image adaptive enhancement arithmetic based on the ratio of image variance to noise
variance which can adjust factor K bases on CR image gray characteristic and enhance image detail and improve image
diagnosis value. Experiment results demonstrate that the algorithm enhances CR image detail and the value of DV/BV
was higher than the other algorithms, moreover CR image enhanced has good visual effect, the method is fit for edge
detail enhancement of CR medicine radiation image.
An implementation of remote sensing images thematic mapping
Mu-hong Dai,
He-qing Li
Show abstract
Remote sensing thematic mapping is an important link in the filed of remote information application. Remote sensing
information has been widely used in more and more departments and fields. Specially, some data are available in
common research. Therefore, it is vital to propose couple of solutions to promote the application of Remote sensing
information. This paper introduced a generalized method and a mass-oriented thematic mapping approach is discussed. It
encapsulates all the intricate steps, so the process became transparent. In the field of remote sensing in water
environment, this method has been widely applied.
A two-stage scheme for multi-view human pose estimation
Show abstract
We present a two-stage scheme integrating voxel reconstruction and human motion tacking. By combining voxel
reconstruction with human motion tracking interactively, our method can work in a cluttered background where perfect
foreground silhouettes are hardly available. For each frame, a silhouette-based 3D volume reconstruction method and
hierarchical tracking algorithm are applied in two stages. In the first stage, coarse reconstruction and tracking results are
obtained, and then the refinement for reconstruction is applied in the second stage. The experimental results demonstrate
our approach is promising. Although our method focuses on the problem of human body voxel reconstruction and
motion tracking in this paper, our scheme can be used to reconstruct voxel data and infer the pose of many specified rigid
and articulated objects.
The vehicle license plate location based on mathematical morphology and geometric characteristics
Guangming Li,
Zhenqi He,
Huilin Zhang
Show abstract
Currently, the problems of low accuracy and slow speed in the LPR (license plate recognition) of motor vehicle
still exist, while the problem is mainly due to complex interference caused by the surrounding environment. As the
collection plate is usually located on the lower portion of the image, this system firstly cut the image from the position of
1 / 2 to obtain the lower part that can effectively reduce the interference information. Then, we orientate license plate
accurately through Canny operator edge detection[1] and the geometric ratio features of license plate[6]. Combining
these two methods can further remove the interference factors, and ultimately obtain accurate positioning. Finally, the
experiment result has proved the method more reliable and faster.
Approach based on visual recognition cylindrical ceramic resistance detection system
Shaobin Ren,
Yuanzong Li
Show abstract
Ceramic substrate is an important substrate of electronic components, the traditional detection methods rely on manually.
Since the product is small size, light weight and a huge amount of testing in the detection process without the stability
and accuracy of the reliable guarantee. Detect small objects at the same time as artificial, easily fatigue. In this paper,
machine vision images for recognition. According to the characteristics of product samples to design a complete
detection system, related to verification by the laboratory. Practical results show that the method be applied to other
products of similar quality small automatic detection.
Distributed joint power and access control algorithm for secondary spectrum sharing
Hongyan Li,
Enqing Chen,
Hongliang Fu
Show abstract
Based on interference temperature model, the problem of efficient secondary spectrum sharing is formulated as a power
optimization problem with some constraints at physical layer. These constraints and optimization objective limit a
feasible power vector set which leads to the need of access control besides power control. In this paper, we consider the
decentralized cognitive radio network scenario where short-term data service is required, and the problem of distributed
joint power and access control is studied to maximize the total secondary system throughput, subject to Quality of
Service (QoS) constraints from individual secondary users and interference temperature limit (ITL) from primary
system. Firstly, a pricing-based game model was used to solve distributed power allocation optimization problem in both
high and low signal to interference noise ratio (SINR) scenarios. Secondly, when not all the secondary links can be
supported with their QoS requirement and ITL, a distributed joint power and access control algorithm was introduced to
find the allowable links which results in maximum network throughput with all the constraints satisfied, and the
convergence performance is tested by simulations.
Research on the testing technology for the water-saving performance of shower nozzle based on image analysis
Show abstract
Spray field of the shower nozzle has its own unique characteristics, which are instantaneity and speediness, gas-liquid
two-phase nature and complexity of multiple jets. Using the traditional artificial observational method to detect the
water-saving performance can not get high precision and can only extract limited characteristics. A method of image
acquisition and analysis technology is described in this article, which can provide foundation for testing water-saving
performance of the nozzle. By the image analysis method, it can obtain both the flow characteristic of the jet in the spray
field and the relationships of pressure, flow rate and spray angle.
3D reconstruction based on a complex constrain matching algorithm
Qiongyan Li,
Feng Ma
Show abstract
The precision of stereo matching directly influence the result of 3D reconstruction from images. a matching method is
proposed to increase matching accuracy. Two pictures are taken for an object in different angle using a digital camera,
then the feature points are extracted from the two images. After that, the initial matching is firstly done for the image
feature points, then accurate matching is performed by use of affine transformation constrain, epipolar geometry
constrain and gray scale relativity constrain comprehensively. Finally the accurate matching points are utilized to
reconstruct the 3D points, which are filtered, smoothed, triangulated and textured to get a true 3D model Experimental
results show that this algorithm converges fast and can increase matching accuracy effectively.
Measuring of optics surface information with digital image processing technology
Shaojun Lu,
Jun Han,
Wang Liang,
et al.
Show abstract
The quality of optics' surface is of an important character which has critical influence in optical system. There are many
ways to detect surface quality, but interferometry is considered as an effective technology. Interfere fringes which
obtained from our experiment based on equal thickness interference were processed with digital image processing (DIP)
technology in this paper. Image smoothing, fringes thinning, fringes' space measurement and P-V value measurement
were done in this processing, Which can give optics' surface information automatically and accurately. We measured
some optics whose apertures are not more than 30mm, and the result is satisfactory. Our method will be widely used in
industrial inspection, especially in optical works in the future.
Study on vision object tracking based on adaptive object segmentation
Hui-Juan Hao,
Ji-Yong Xu,
Guang-Qi Liu
Show abstract
In order to solve the object tracking under occlusion, the adaptively tracking algorithm is proposed based on color
features. The object is adaptively divided using fuzzy k-means clustering algorithm, and the sub-regions are weighted
with monotone decreasing kernel function. The object model is updated through mean value of sub- regions' colors, so
the calculation is simple. During the object tracking, the method of integral matching is used; combining with the
adaptive Kalman filter, the object tracking under occlusion is resolved effectively. The experiments show that the new
algorithm can track the object exactly.
Fruit shape classification using Zernike moments
Show abstract
A new method along with Zernike moments for classify fruit shape is developed, the image is first subjected to a
normalization process using its regular moments to obtain scale and translation invariance, the rotation invariant Zernike
features are then extracted from the scale and translation normalized images and the numbers of features are decided by
primary component analysis (PCA), at last, these features are input to support vector machine (SVM) classifier. This
method performs better than traditional approaches because of their orthogonal base and rotation invariance of the
defined features on them, which is verified by experiments on Zernike moments and Fourier descriptors.
Sort entropy-based for the analysis of EEG during anesthesia
Liang Ma,
Wei-Zhi Huang
Show abstract
The monitoring of anesthetic depth is an absolutely necessary procedure in the process of surgical operation. To
judge and control the depth of anesthesia has become a clinical issue which should be resolved urgently. EEG collected
wiil be processed by sort entrop in this paper. Signal response of the surface of the cerebral cortex is determined for
different stages of patients in the course of anesthesia. EEG is simulated and analyzed through the fast algorithm of sort
entropy. The results show that discipline of phasic changes for EEG is very detected accurately,and it has better noise
immunity in detecting the EEG anaesthetized than approximate entropy. In conclusion,the computing of Sort entropy
algorithm requires shorter time. It has high efficiency and strong anti-interference.
The image subpixel-precise measurement of holes in mechanical part based on computer vision
Show abstract
The measurement method of mechanical part based on computer vision is a new trend with high accuracy, high
speed and automation, which plays a very important role in machining industry. In this paper, a new measurement
method of circular hole of mechanical part is proposed based on the subpixel-precise thresholding and contour
extraction. Under the base of the new measurement method, a computer vision system is designed to measure the
position of hole, diameter of hole, circularity of hole, and the distance between relevant two holes with high speed,
high efficiency and subpixel-precise. Good results have been got in the measurement experiment, which proves the
validity of the new measurement method.
Morphological color image processing based on improved distance and lexicographical ordering
Hongzhong Tang,
Huixian Huang,
Yewei Xiao,
et al.
Show abstract
The paper presents a novel definition method of color morphological operators. It is followed the combination of
weighted distance ordering based on the reference color and improved lexicographical ordering. L*a*b* color model is
integrated with its polar coordinates representation, weighted impact factors are employed, vector ordering is initially
realized by mapped to scalar function. The lack of equidistant is solved with the lexicographical ordering, the improved
lexicographical ordering is introduced based on rough set. It is possible to reduce the first component with attributions
and contributions of the preceding components in ordering. By comparative application to color image processing,
experimental results demonstrate favorable consistency in terms of human perception and better image color balance,
details and performance of image show more richer and smoother.
Feature matching method in shaped light mode VFD defect detection
Show abstract
In recent years, Vacuum Fluorescent Display (VFD) module in the car audio panel has been widely used.
However, due to process reasons, VFD display production process will produce defects, not only affect the
appearance, but also affect the display correctly. So building a car VFD display panel defect detection system is of
great significance. Machine vision technology is introduced into the automotive VFD display defect detection in
order to achieve fast and accurate detection of defects. Shaped light mode is a typical flaw detection mode which is
based on characteristics of vehicle VFD panel. According to the image features, learning of the gray matching and
feature matching method, we integrated use of feature matching method and the gray level matching method to
achieve defect detection.
Research of spatial high-pass filtering algorithm in particles real-time measurement system
Show abstract
With the application development of CIMS, enterprises have the more need of the CAQ systems during the
process of flexibility and automation. Based the means of computer-based vision technology, Automated Visual
Inspection (AVI) is a non-contact measurement mean synthesizing the technologies such as image processing,
precision measurement. The particles real-time measurement system is the system which analyzes the target image
obtained by the computer vision system and gets the useful measure information. In accordance with existing prior
knowledge, the user can timely take some measures to reduce the floating ash. According to the analysis of the
particle images, this paper researches the image high-pass filter means, Gradient arithmetic, with characteristics of
images. In order to get rid of the interference of background and enhance the edge lines of particles, it uses the two
directions kernel to process the images. This Spatial high-pass filtering algorithm also helps to conduct the ensuing
image processing to obtain useful information of floating ash particles.
Merging algorithm of vector graphics primitive based on reliability index for engineering drawings
Bin Cheng,
Shusheng Zhang,
Yunfei Shi
Show abstract
Vectorizing recognition for engineering drawings is currently one of the most active research topics in the domain of
pattern recognition. The merging algorithm of vector graphics primitive based on reliability index is presented.
According to the dimension of vector basic graphics primitive, the reliability index of vector basic graphics primitive is
defined, and then the reliability index of vector graphics primitive is set. Based on the reliability index, the vector
graphic primitive is divided into two groups. One is the vector graphics primitive with high reliability index, the other is
the vector graphics primitive with low reliability index. Firstly, Taking the graphics primitive with high reliability index
as seed graphics primitive to carry out the merging algorithm, when searching the all vector basic graphics primitive is
completed, and then the second merging algorithm is produced taking the vector graphics primitive with high reliability
index as extending seed until the all vector graphics primitive is searched completely. The algorithm is developed based
on the engineering drawing recognition system (EDRS) to merge the basic graphics primitive obtained by EDRS. The
algorithm presents good results to merge the vector basic graphics primitive.
Recognition conversion of circular graphic object for engineering drawings: the state of the art and challenges
Bin Cheng,
GuangMing Lei,
YaHong Wang
Show abstract
This strong interest to vectorization and recognition for engineering drawings is driven by a wide spectrum of promising
applications in many areas such as 3D reconstruction, information interchange, content-based retrieval to engineering
drawings, 2D understanding for engineering drawings and so on. This paper reviews development level and recognition
methods of vectorization and recognition of circular-typed graphics object for engineering drawings, and provides a
comprehensive survey of research history and the state of the art in recognition conversion of circular-typed graphics
object for engineering drawings. The approach of conversion circular-typed graphics object are divided into fitting-based
methods, vector-based methods, global-based methods and integrated-based methods. The advantage and disadvantage
of all existing model of circular graphics object are analyzed and compared. Research difficulties and development
tendency in future are clearly discussed for engineering drawings. At the end of this survey, some detailed discussions on
research challenges and future directions of vectorization and recognition for engineering drawings are also provided.
Framelet lifting in image processing
Da-yong Lu,
Tie-yong Feng
Show abstract
To obtain appropriate framelets in image processing, we often need to lift existing framelets. For this purpose the paper
presents some methods which allow us to modify existing framelets or filters to construct new ones. The relationships of
matrices and their eigenvalues which be used in lifting schemes show that the frame bounds of the lifted wavelet frames
are optimal. Moreover, the examples given in Section 4 indicate that the lifted framelets can play the roles of some
operators such as the weighted average operator, the Sobel operator and the Laplacian operator, which operators are
often used in edge detection and motion estimation applications.
3D face detection and face recognition: state of the art and trends
Xilai Li,
Aihua Li,
Xiangfeng Bai
Show abstract
Face detection and recognition are old but rapidly growing research areas due to increasing demands for security in
industrial and law enforcement applications. Face recognition based on two-dimensional (2D) images have reached a
significant level of maturity with some practical success. However the performance of 2D face recognition may degrade
under poor illumination conditions, pose changing or for subjects of various skin colors. 3D face recognition represents a
viable alternative to 2D face recognition in the research for a robust and practical identification system. This paper
provides a review of research achievement in face recognition from 2007 to present. This survey mainly include three
parts: (1) advances in face detections, (2) state of the art of 3D face recognition and (3) face images database and face
recognition algorithm evaluation. Recent research has also demonstrated that the fusion of 2D and 3D face recognition
can improve the overall performance of face recognition algorithm.
A new video surveillance system based on AVS
Tingwan Wu,
Chaofeng Chen
Show abstract
In this paper, we analyzed the key technology of current video surveillance systems, compared the performances, prices,
and the prospects between the H.264 and AVS codec technologies, studied the advantages and disadvantages of different
designs, particularly the data transmission, terminal reception facilities, decoding and demonstration of the video
surveillance systems. We have optimized the core algorithms such as intra prediction, inter prediction and fast motion
estimation, and performed some relevant simulations for these algorithms. Experimental results show that these new
algorithms simplified the design and speeded up the compression rate. From the simulation results, we proposed a new
video surveillance system based on AVS. The program reached low price, high efficiency, fast, high image quality,
convenient application and operation, greatly improved the current video surveillance systems.
Research on smoothing support vector regression based on cubic spline interpolation
Bin Ren,
LiangLun Cheng
Show abstract
Smoothing functions can transform the unsmooth support vector regression into smooth ones,and thus better
regression results are generated. In the paper,using Cubic Spline Interpolation, a new polynomial smoothing function is
proposed for the [equation] function in ε-insensitive support vector regressions. Theoretical analysis shows that
[equation] function is better than [equation]-function in properties, and the approximation accuracy of the proposed smoothing function is two order of magnitude higher than that of the [equation]
-function. The experimental results show the validity of the
model. Therefore, the new better polynomial smooth functions is provided for smoothing support vector regression and
related areas.
3D face recognition for security and defense engineering
Xilai Li,
Aihua Li,
Xiangfeng Bai
Show abstract
Face recognition based on 2D images is a difficult problem, because the illumination, pose and expression changes in the
images create great statistical differences and the identity of the face itself becomes shadowed by these factors. 3D face
recognition has the potential to overcome feature localization, pose and illumination problems, and it can be used in
conjunction with 2D systems. We review the relevant work on 3D face recognition here, and discuss merits and demerits
of different representations and recognition algorithms. This survey focuses on 3D face data acquisition and 3D face
recognition using histogram features. Future works involved in developing more accurate 3D face recognition are also
discussed. These include the need for improved 3D face data acquisition, 3D face recognition algorithms and 3D face
experimental methodology.
An algorithm for subpixel edge location of circular features based on Zernike moments and moment-preserving
Qingbin Tong,
Xiaodong Zhang,
Zhenliang Ding,
et al.
Show abstract
In order to satisfy the stringent requirements for the high accurate detection of circular feature in such applications as the
vision measurement of electronic components and camera calibration, based on the principle of orthogonal Zernike
moments and simple moment-preserving transform, a subpiexl edge detection algorithm for the curve of circular feasures
is proposed. A circular-arc geometry is assumed for the boundary inside the detection area of template. The new arc-edge
detector is designed as a cascade process using Zernike linear-edge detector and simple moment-preserving transform
and a look-up table. The validity and detecting precision for the algorithm are studied. The experimental results show
that the algorithm is very accurate and stable, and the measuring accuracy is better than 0.03 pixel. The proposed method
is very effective in detecting circular features and as well ellipse curve for vision measurements.
Study on the application of MRF and fuzzy clustering as well as the D-S theory to image segmentation of the human brain
Show abstract
A new image segmentation method based on Markov Random Field (MRF) and Two-Dimensional Histogram Method of
Fuzzy Clustering as well as Dempster-Shafer (D-S) evidence theory is presented in this paper.The application of Markov
Random Field to image restoration and segmentation can effectively remove noise and get more accurate segmentation
results; And the application of Fuzzy Clustering Theory together with Two-Dimensional Histogram image segmentation
methods can get more satisfactory segmentation results; However, these two ways leads to different classification results
while classifying the controversial pixels in images, so we can use the Dempster-Shafer evidence theory to assign the
controversial points to the plausibility interval, and then divide them. This paper will adopt the above three theories to
propose a human brain image segmentation research method. Experimental result shows that the method solves the
problem of the class attribution of the controversial points, and the segmentation result is more in line with human vision.
A calibration method of monocular stereo vision system for on-machine measurement
Show abstract
A calibration method of movable monocular stereo vision measurement system for on-machine measurement was
presented based on standard dimension. The physical model and mathematical model was established in view of
perspective projection transformation. Structure parameters of this stereo vision system were determined using
Levenberg-Marquardt algorithm. Experiment results show that this calibration method can achieve precise camera
parameters and three dimensional space coordinates and can be employed conveniently in real industrial environment.
A new algorithm for image denoising based on tetrolet transform
Cai-lian Li,
Ji-xiang Sun,
Yao-hong Kang
Show abstract
This paper introduces a new class of denoising function that has continuous derivative for image denoising. And a
new algorithm are presented. First, we apply tetrolet transform to noise image and obtained tetrolet coefficient. Second,
by using the new denoising function, we present an adaptive method based on SURE Risk. Instead of the global
hard-thresholding algorithm for image denoising, we minimize an estimate of the mean square error by using adaptive
genetic algorithm. At last Numerical experiments show that the proposed new algorithm can significantly outperform
the original hard-thresholding method both in terms of PSNR and in visual quality.
Generation of geographical profile
Zhi-Bin Shen,
Yuan-Biao Zhang,
Kai-Fa Liang,
et al.
Show abstract
To provide help for the police's investigation on serial criminals, we develop a mathematical model in the
paper. First, we use Inherently Continuous Model and Improved Kinetic Model to generate the offender's
geographical profile. However, there is a difference in two models' results. For better synthesizing the difference,
we develop a Combination Model and generate a new geographical profile. As a result, we estimate the offender's
location and carry on a series of analysis. What's more, the models created can be applied in other fields, such as
market's investigation, military operations and so on.
Image deblurring with adaptive total variation model
Yang Bai,
Yuanyuan Ding,
Xin Zhang,
et al.
Show abstract
In this paper the models of blurring and methods of deblurring are introduced. A kind of nonlinear image deblurring
approach is discussed, which comes from constrained optimization total variation approaches. An adaptive TV
model method of image deblurring based on common TV model methods is proposed. The experimental results
indicate that the proposed method can protect the details of the blurry image more efficiently and satisfy some
deblurring requirements more adaptively.
Application of support vector machine and quantum genetic algorithm in infrared target recognition
Hongliang Wang,
Yangwen Huang,
Haifei Ding
Show abstract
In this paper, a kind of classifier based on support vector machine (SVM) is designed for infrared target recognition. In
allusion to the problem how to choose kernel parameter and error penalty factor, quantum genetic algorithm (QGA) is
used to optimize the parameters of SVM model, it overcomes the shortcoming of determining its parameters after trial
and error in the past. Classification experiments of infrared target features extracted by this method show that the
convergence speed is fast and the rate of accurate recognition is high.
A new edge-based interactive image segmentation method
Qiang Chen,
Quan-sen Sun,
De-shen Xia
Show abstract
This paper presents a new edge-based interactive image segmentation method. First an edge image is generated with an
edge detector, and then primary object edges can be selected by separating the object and background edges. Finally we
obtain the complete object edge by connecting broken object edges. In order to reduce the influence of the noise, image
denoising can be done before the edge detecting. Natural and medical image segmentation demonstrates that our method
incorporates the advantages of the automatic and manual image segmentation methods. For most of images, a little of
manual object edge connection is needed to obtain object edges. It is an effective and feasible interactive image
segmentation method.
Print image sharpness analysis based on gray-level co-occurrence matrices
Lin Zhang,
Meiyun Zhang,
Yangyu Wu
Show abstract
A novel measure is presented to quantify print image sharpness. Nine texture features of gray level co-occurrence
matrices (GLCM) were calculated from the print images respectively which were blurred by Gaussian blurs filter with
different radius ranging from 0 to 8 pixels in steps of 2. Experiments were performed on these images with different
GLCM distance d (2, 4, 6, 8,10 pixels) and orientation θ (0°, 45°, 90°, 135°) under the constant window size (64 pixels).
Furthermore, the correlation matrix of texture features was calculated to judge which texture features can be chosen to
assess sharpness most. The test results show contrast and energy provide the most unique information of print image
sharpness. And the distance d of GLCM can be determined to be 6 pixels and the different orientation θ has little effect
on the trends. The method is reliable and extends GLCM with the sharpness evaluation of variable size, oriented print
image.
A fusion method of metallurgical images based on curvelet transform
Wencheng Wang,
Faliang Chang,
Lei Wang
Show abstract
On the basis of analyzing several common algorithms of image fusion, a new multi-focus image fusion method based on
curvelet transform is proposed according to the question of metallurgical image fusion. Firstly, two different focal
images were decomposed using curvelet transform respectively, and then in the curvelet domain of the two transformed
images, the new curvelet coefficients were acquired by adopting a simple fusion rule, which was that the low-frequency
coefficients were integrated using the weighted average, and high-frequency coefficients were integrated using choose
max. Finally, the fused coefficients are reconstructed to obtain fusion results. Experimental results show that the method
is more suitable for metallurgical image fusion than other ways, the fused image will have more information and the
observation with naked eyes will be clearer.
Application of filtering techniques in preprocessing magnetic data
Haijun Liu,
Yongping Yi,
Hongxia Yang,
et al.
Show abstract
High precision magnetic exploration is a popular geophysical technique for its simplicity and its effectiveness. The
explanation in high precision magnetic exploration is always a difficulty because of the existence of noise and
disturbance factors, so it is necessary to find an effective preprocessing method to get rid of the affection of interference
factors before further processing. The common way to do this work is by filtering. There are many kinds of filtering
methods. In this paper we introduced in detail three popular kinds of filtering techniques including regularized filtering
technique, sliding averages filtering technique, compensation smoothing filtering technique. Then we designed the work
flow of filtering program based on these techniques and realized it with the help of DELPHI. To check it we applied it to
preprocess magnetic data of a certain place in China. Comparing the initial contour map with the filtered contour map,
we can see clearly the perfect effect our program. The contour map processed by our program is very smooth and the
high frequency parts of data are disappeared. After filtering, we separated useful signals and noisy signals, minor
anomaly and major anomaly, local anomaly and regional anomaly. It made us easily to focus on the useful information.
Our program can be used to preprocess magnetic data. The results showed the effectiveness of our program.
Automatic pipes counting system based on digital image processing technology
Wencheng Wang
Show abstract
In order to improve the efficiency and accuracy of pipes counting, a system was designed based on the machine vision
and image processing technology in this paper, which can automatically count the number of the pipes online. Firstly, the
image was captured by CCD and input into computer for processing. Then, a kind of gray image preprocessing methods
such as denoising, binary conversion, mathematical morphology have been conducted. Finally, an accurate counting
strategy of pipe image is adopted based on 8-neighborhood region labeling algorithm. Experiment results show that this
method is convenient, and it is effective to deal with the conglutination problems and enhances the accuracy of the pipe
automatic counting.
Evolutionary extreme learning machine based on dynamic Adaboost ensemble
Gaitang Wang,
Ping Li
Show abstract
Boosting ensemble algorithm exhibits two fatal limitations: one is that it gives in advance the upper bound of weighted
error on weak learning algorithm; the other one is that it is overdependent on data and weak learning machine, and it is
too sensitive to data noising. Aimed at limitation of Boosting ensemble application in extreme learning machine, this
paper proposes a new algorithm: evolutionary extreme learning machine based on dynamic Adaboost ensemble, which
regards the evolutionary extreme learning machine as weak learning machine, dynamic Adaboost ensemble algorithm is
used to integrate the outputs of weak learning machines, and makes use of fuzzy activation function as activation
function of evolutionary extreme learning machine because of low computational burden and easy implementation in
hardware. Proposed algorithm has been successfully applied to problem of function approximation and classification
application. Experimental results show that the algorithm increases the training speed greatly when dealing with large
dataset and has better generalization performance compared to extreme learning machine, evolutionary extreme learning
machine and Boosting ensemble extreme learning machine with quasi-Newton algorithms.
Number extraction in document images based on mathematical morphology
Show abstract
Number extraction in document image is a crucial technique in document image analysis. To simply and efficiently
extract the numbers, a mathematical morphology based algorithm is proposed in this paper. Firstly, the square regions
containing numbers are labeled by morphological dilation operation using a designed structuring element. Secondly, the
square regions are extracted by morphological erosion operation. Thirdly, the inner region of the square regions is
extracted through morphological dilation operations. Finally, the numbers are extracted through comparing the extracted
and original inner region of the square regions. Experimental results show that the proposed algorithm is efficient.
Improvement mean shift-based image segmentation approach for automatic agriculture vehicle
Yong-hua Han,
Ya-ming Wang,
Yun Zhao
Show abstract
Mean Shift algorithm, a statistic iterative procedure, is robust when applied to farmland image segmentation. It
can effectively overcome the influence of shadow, weeds or illumination changes, etc. However, the Mean Shift
procedure has relatively high time complexity and can not meet the requirements of real-time processing. Based on
pyramid algorithm, we can obtain a low resolution representation of the images being processed. Then, run Mean shift
algorithm on a set of seed points that selected in the low resolution image. Through this method, the time consumption is
significantly lower than the original Mean Shift Procedure. The objects in farmland images are large and there are only
two major types of structure in it, so the examination accuracy of proposed method is changed little. At the same time
based on spatial structure and color distribution of farmland image, Mean Shift Kernel radius in the spatial and range
domain is selected. In addition, according to different seasons, crops show different colors. In this case, the equations
which convert color image into a grayscale image are discussed.
A surface fractal dimension research of soil-pore interface based on image analysis
Show abstract
The inner spatial formation and structure of soils has been the basic and important subject of investigation. With the
development of digital image technology, it's easier to obtain images of soil-pore at the extent between micron and
millimeter. In this study, we acquire soil-pore binary image by application of Fast Fourier Transform (FFT) and filter
technique, combined with actual soil porosity. The soil-pore fractal dimension depicts the irregular and rugged
boundaries between soil pores and particles. The fractal dimension research of soil-pore interface indicates the obvious
correlativity existed between fractal properties and soil texture, though pore-size distribution pattern significantly alters
soil properties.
Situation estimation modeling and simulation of aerial optical-electronic sensors
Zhi-fu Shi,
Hai-yan Liu
Show abstract
Aim at the information of aerial optical-electronic sensors, the multi-phase information fusion model was built for
situation estimation. The method first builds a data association of target flight path with fuzzy cluster based on aerial
optical-electronic sensors' information. The data fusion of target information is done next via maximum likelihood
estimation on cluster attribute. Then the uniform expression for target feature was gained. These are then fed into fuzzy
Bayesian networks that perform inference via belief propagation for situation estimation. The inference result can
provide decision-maker with technique foundation. A simulation example of the whole situation estimation process
demonstrates the validity of the model and the reliability of the inference results.
An improved fusion of IHS based on wavelet for high resolution images
Yi Feng,
Beiping Wu,
Yingchun Yue
Show abstract
In recent years, Multi-sources data fusion techniques have already been an International research hotspot in Remote
Sensing. To date, many image fusion techniques have been developed. However, the available algorithms can hardly
produce a satisfactory fusion result for high resolution images. Among the existing fusion algorithms, the IHS technique
is the most widely used one, and the wavelet fusion is the most frequently discussed one in recent publications because
of its advantages over other fusion techniques. But the shortcome of colour distortion and low resolution in many field is
often obvious, especially when QuickBird natural colour multispectral images are fused with its panchromatic images. In
this paper, an improved fusion of IHS based on wavelet is proposed. At the same time ,From the result of this experiment
proves that the concept of the proposed improved fusion is promising, and it does significantly improve the fusion
quality compared to conventional IHS and wavelet transform fusion techniques.
Research and development of synchronous controller in the three-dimensional measurement of a moving object
Jianfu Chen,
Genyuan Zhou,
Hui Zhang,
et al.
Show abstract
By means of analyzing characteristics of three-dimensional measurement of moving objects, a method of data collection
based on active optical three-dimensional measurement is proposed. At the same time, to overcome shortcomings of
classical synchronization method, the external trigger-type synchronous means and the dimming and the temperature
measurement of pulse width modulation (PWM) technique are proposed. The synchronous controller is developed with
the core of MCU AT89C2051 to ensure the synchronous of every camera and projector. Finally, An application case of
prototype system validated the effectiveness of synchronization accuracy in the image acquisition system. It shows that
the common solution of synchronization is provided in three-dimensional measurement of moving objects, flow rate and
flow measurement.
Assessment method to fusion effect based on structural similarity comparison in fusion images
Show abstract
Image fusion can integrate several images of the same scene captured by several different sensors with different features
and resolutions at different time into one image. Research on quality assessment of fusion images is meaningful for
image processing course in order to improve the registration technology and fusion algorithm. Structural similarity
metric describes differences between two images by means of three variables, luminance, contrast, and spatial similarity,
which show the better evaluating capability than others objective metrics. A new assessment method to fusion effect
based on structural similarity comparison among fusion images is provided in paper. Fusion algorithms including
weighing method, principal component analysis, different pyramid methods and multi-resolution wavelet filtering is used
to create fusion images. Then the mutual structural similarity metric among fusion images obtained by different fusion
algorithms is used to evaluate the fusion effect. In some extent, the low structural similarity comparison denotes the low
quality fusion effect. Meanwhile, the experiment show also the fusion effect determined by structural similarity
comparison is accordant with the subjective evaluation. Besides, the experiment explain the method based on different
pyramid methods and multi-resolution wavelet filtering have the better fusion effect than weighing method and principal
component analysis method. Furthermore, the experiment also prove the whole image fusion system should choose the
different fusion algorithm to adjust to the different task requirement and applied circumstance in order to acquire the
optimum scene interpreting effect.
Photogrammetric measurement of the ASKAP antenna
Qi-qiang Feng,
Zong-chun Li,
Guang-yun Li
Show abstract
The Australian SKA Pathfinder ratio telescope (ASKAP) comprises an array of up to 45 dish antennas with diameter of
12 meters. ASKAP is situated at Australia's candidate site for the International Square Kilometre Array telescope in
Western Australia. It is intended to be a fully-functional radio telescope, with a science case based on its wide-field-of-view
capabilities at centimeter wavelengths. The paper reports the photogrammetric measurement of the first ASKAP
antenna reflector surface. Deformation of the reflector surface was measured with photogrammetric system V-STARS at
six different elevation and polarization angles. The maximum RMS of orthogonal distance residuals from the best-fit
paraboloid is 0.58mm, less than the design accuracy of 1mm.
A CT image matching algorithm based on singular value decomposition
Ping Han,
Jing-fei Sun,
Hui-hui Mou
Show abstract
To strengthen the noise immunity of images matching, and enhance the robustness of matching, we advanced an image
matching algorithm based on singular value decomposition. Firstly, it fetched the feature point of reference drawing and
floating drawing by using Harris Algorithm, and then matched the feature point set of the two drawings with singular
value decomposition method. Because of the Orthogonal and similar properties of singular value decomposition, the
noise immunity of this algorithm was stronger, and the robustness of matching was higher. The experimental result
proves the effectiveness of this algorithm.
Locomotive wheel 3D reconstruction
Show abstract
In the article, a system, which is used to reconstruct locomotive wheels, is described, helping workers detect the
condition of a wheel through a direct view. The system consists of a line laser, a 2D camera, and a computer. We use 2D
camera to capture the line-laser light reflected by the object, a wheel, and then compute the final coordinates of the
structured light. Finally, using Matlab programming language, we transform the coordinate of points to a smooth surface
and illustrate the 3D view of the wheel. The article also proposes the system structure, processing steps and methods, and
sets up an experimental platform to verify the design proposal. We verify the feasibility of the whole process, and
analyze the results comparing to standard date. The test results show that this system can work well, and has a high
accuracy on the reconstruction. And because there is still no such application working in railway industries, so that it has
practical value in railway inspection system.
The application of pattern recognition in wood processing industry
YeQin Wang,
Hui Wang
Show abstract
In order to improve the level of automation in wood production, the parameters of Gray level co-occurrence
matrix(GLCM) and Gauss - Markov random field(GMRF) were extracted. S - NFS algorithm was applied to data fusion.
And the redundancy and complementarities of two texture parameters were used to build the wood texture parameter
system. An integrated measurement rule based on BP neural network classifier's overall recognition rate of samples was
advanced to design its integrated classifier. Experiments show that the recognition rate of integrated neural network
classifier is superior to the individual network and the nearby classifier, and the average recognition rate of 10 texture
samples have reached up to 97%, which could meet the needs of industrial production. And it shows that the established
parameter system for wood texture description is effective.
Bundle of paper money recognition based on image entropy
Limin Wang,
Yanxun Wang,
Hongliang Zhang,
et al.
Show abstract
In order to consummate the management of bank paper, a new algorithm which can recognize the areas of paper tape in
bundle of paper money is presented. Image segmentation is used to divide the image into several sub-blocks.
Specifically, image entropy is computed in each sub-block. Experimental in different algorithms, including color training
algorithm and the median filter and edge detection algorithm, demonstrates its real-time, effectiveness and robustness.
Fast enhanced face-based adaptive skin color model
Chen-Chiung Hsieh,
Dung-Hua Liou,
Meng-Kai Jiang
Show abstract
Man machine interface by video analysis becomes popular recently. The most typical body gesture utilized for computer
interaction is hand gesture. Therefore, it is a very important topic to accurately extract hand regions from a sequence of
images in real time. In this paper, we propose an adaptive skin color model which is based on detected face color. Skin
colors are sampled from extracted face region where non-skin color pixels like eyebrow or glasses are excluded.
Gaussian distributions of normalized RGB are then used to define the skin color model for the detected person. To
demonstrate the robustness of proposed model, experiments under diversified lighting and background are tested.
Traditional methods based on RGB, Normalized RGB, and YCbCr are all implemented for comparison. From
experimental results, skin color pixels could be detected for each person. The accuracy rate is 95.73% on average and is
superior to previously mentioned methods.
A fuzzy binarization method for complicated background document images
Yuyan Chao,
Lifeng He,
Kazuhito Murakami,
et al.
Show abstract
This paper presents a fuzzy binarization method for document images with complicated backgrounds. Our binarization
method consists of a preprocessing procedure, a character-line detector, and three binarization processings. The first
binarization processing is used to obtain a basic binary image, the second is made by considering edge information, and
the third does by considering background information. Experimental results demonstrate that our fuzzy binarization
method is efficient for separation of foreground (text) areas and background areas in images with stain areas or
complicated backgrounds.
An efficient run-based connected-component labeling algorithm for three-dimensional binary images
Lifeng He,
Yuyan Chao,
Kenji Suzuki,
et al.
Show abstract
This paper presents an run-based efficient label-equivalence-based connected-component labeling algorithms for threedimensional
binary images. Our algorithm is run-based. Instead of assigning each foreground voxel, we assign each run a
provisional label. Moreover, we also use run data to label foreground voxels without scanning any background voxel in
the second scan. Experimental results have demonstrated that our algorithm is much more efficient than conventional
three-dimensional labeling algorithms.
A global registration method for temporal subtraction of chest radiographs
Lifeng He,
Takeshi Inaba,
Kenji Suzuki,
et al.
Show abstract
This paper presents a mutual-entropy-based global registration method for temporal subtraction of chest radiographs.
Our global registration method consists of two steps. In the first step, we make a rough registration by calculating the
mutual entropy of on the whole areas of the previous radiograph and the current radiograph, and in the second step, we
make an accurate registration by calculating the mutual entropy of on the rib areas only. Experimental results
demonstrated that the visualization of pathological change in the temporal subtraction images constructed by our global
registration method is better than that constructed by convenient global registration methods.
A substation infrared temperature monitoring and warning system with object separation and image registration
Lihua Lin,
Dongmei Wu,
Xinghua Zhang
Show abstract
To find the defects of the apparatus in a substation in the early stage, an infrared temperature monitoring and
warning system is established. This system can monitor the electrical equipment automatically the movement condition.
The systemic circulation gathers the transformer substation electrical equipment the infrared imagery, the extraction goal
equipment temperature information, and with the history database creation connection, the synthesis distinguishes the
equipment failure information. In view of image gathering when because the mechanical drive creates the deviation,
proposed one kind of object-oriented division and the image matching adjustment algorithm, first carries on the object
division and the configuration definition to the image, then uses based on the phase correlation carries on the matching
with the Harris vertex match image matching method to the deviation image. In this paper, a infrared remote-viewing
image registration based on phase correlation and feature points matching is presented. Several experiments illustrate
that this method has a good performance of reliability and accuracy.
Globally consistent image mosaicing
Show abstract
Image mosaicing is widely used in computer vision applications. Accurate and consistent alignment of sequence images
is the key issue to image mosaicing. In this paper, a globally consistent image mosaicing is proposed by taking account
of various uncertainties. The problem of global alignment of a sequence of images is considered as a stochastic
estimation problem. The transformation parameters of images are considered as system state. System augmentation
model and system observation model are constructed. The global homographies parameters of sequence images are
estimated recursively with augmented Kalman filter in a common state vector and covariance matrix. The proposed
image alignment method can handle the uncertainty efficiently and is globally consistent. Some experimental results are
provided to validate the performance of the proposed method.
Image segmentation based on data field and cloud model
Kun Qin,
Leihai Ou,
Tao Wu,
et al.
Show abstract
There are many uncertainties in image segmentation, which needs theories and methods with uncertainty to handle. This
paper proposes a novel method of image segmentation based on data field and cloud model, which considers the spatial
information of image through data field, and handles the uncertainty of image through cloud model. The proposed
method inspired from cognitive physics considers each pixel as a physical object, calculates the interactive force of these
physical objects, and generates image data field and the potential values which are considered as spatial information.
And then, uses cloud transformation and magnitude cloud synthesis to extract the concepts of potential-frequency
histogram from low level to high level, realizes the clustering of pixels, finally uses maximum determination to partition
the pixels into different classes and segment image into different regions. Results of many experiments indicate that the
proposed method obtains better effect than those of Fuzzy C-means clustering, Otsu and cloud based hierarchical method,
and it is feasible and effective.
A noise filtration technique for fabric defects image using curvelet transform domain filters
Show abstract
A noise filtration technique for fabric defects image using curvelet transform domain Filters is proposed in this paper.
Firstly, we used FDCT_WARPING to decompose image into five scales curvelet coefficients. Secondly, the proposed
algorithm distinguished major edges from noise background at the third scale. Thirdly, the possible lost edges in the
procedure above were detected according to the decaying lever of the coefficients. Fourthly, the edges of the defect at the
second scale were detected by four correlation coefficients in the two directions at the third scale. Fifthly, the curvelet
coefficients at the fourth scale are filtered by the decaying lever. Sixthly, the curvelet coefficients at the fifth scale are
filtered by hard threshing. Finally, the processed coefficients are reconstructed. The tests on the developed algorithms
were performed with images from TILDA's Textile Texture Database, and suggest that the new approach outperforms
wavelet methods in image denoising.
A quick scan and lane recognition algorithm based on positional distribution and edge features
Jian Wang,
Yuan Zhang,
Xiaomin Chen,
et al.
Show abstract
With the growing number of vehicles on the road, the automatic guided vehicles (AGV) vision system for intelligent
vehicles has been given more and more attention. Lane recognition is an important component in the automatic guided
vehicles (AGV) vision system for intelligent vehicles. To improve the speed and accuracy of lane recognition, this paper
proposed an image segmentation algorithm based on the normalized histogram matching and a specific image scan
algorithm based on positional distribution of lanes to reduce runtime. The purpose of image segmentation is extracting
useful road information and the algorithm is segmenting the image by calculating the similarity of Cumulative
Distribution Function (CDF) of normalization histogram. The main idea of image scan algorithm proposed in this paper
is regarding the lanes that have been found as starting points and looking for the new lanes. Then we use a novel lane
screen algorithm based on the left and right edges of lanes' geometric feature to remove invalid information and improve
the accuracy and promote efficiency effectively. At last, a lane prediction algorithm is proposed to predict the farther
lanes which may be lost due to treating as noises. After our tests, this algorithm has better robustness and higher
efficiency.
Region-based document image denoising
Qing-Wen Zhou,
Kai Wang,
Hong-Jiang You,
et al.
Show abstract
Traditional image de-noising methods mainly focus on the global effect, but the noise in document images tends to
gather in certain local areas. So global de-noising processes will inevitably affect the recognition rate. Region based
image de-noising uses pixel statistical information of local regions to separate noise and non-noise regions. And de-noising
only applies on these noise regions instead of the entire image. So the deficiency of traditional method can be
overcome. Test result on UNLV with 11176 samples shows that the average recognition rate rises from 94.44% to
94.85% by using this method.
Research on the consistency of LVQ classifier
Qing-Wen Zhou,
Kai Wang,
Qing-Ren Wang
Show abstract
As a self-organizing artificial neural network model based on supervised learning, the LVQ classifier has been widely
applied and deeply studied due to its good practical performance on the pattern recognition problems. The improved
LVQ classifier have been greatly developed in previous works, and the experimental results on specific problems show
that the improved LVQ classifier is indeed better than the standard learning algorithms proposed by Kohonen. Different
from previous works, the consistency is studied in this paper to provide a theoretical support for the LVQ classifier.
Furthermore, a simulation is included in this paper to provide an experimental support for our theoretical work.
Cryptography based on spatial chaotic system
Fuyan Sun,
Zongwang Lü
Show abstract
Encryption of images is different from that of texts due to some intrinsic features of images such as bulk data capacity
and high redundancy, which is generally difficult to handle by traditional methods. This paper proposes a new spatial
chaos system(SCS), which is investigated by conducting FIPS 140-1 statistic test, and is especially useful for encryption
of digital images. It is shown how to adapt a two dimensional(2D) ergodic matrix obtained from SCS to permute the
positions of image pixels and confuse the relationship between the cipher image and plain image simultaneously.
Experimental results show that the performance and security of the proposed cryptographic system are better than those
of existing lower dimensional chaotic cryptographic systems.
An algorithm of multi-structure based on Riemannian manifold learning
Show abstract
Riemannian Manifold Learning (RML) is a global algorithm proposed recently, so it can't preserve the local geometry
property of neighboring data well. An algorithm of multi-structure based on RML is proposed in order to solve the
problem. In the algorithm, all points were projected by PCA firstly so as to the extracted character is irrelevant, then
constructed a neighbor graph. The most important step was that all data points were classified to two parts, for the
k - NN of a base point, it adopted a weight which can preserve local property of the base point and neighboring nods to
get the low-dimensional embedding coordinates. As for the other points, it still used the RML algorithm. Thus the new
algorithm can both preserve the metrics at all scales and keep the geometrical property of local neighbor to the
maximum. Experimental results on synthetic data and MNIST data set demonstrate that the new algorithm can reflect the
intrinsic property better than the other manifold learning algorithms.
Application of pattern recognition in the forecast of outburst area of coal and gas
Show abstract
Based on the intrinsic relation between a number of outburst factors and outburst dangers, multi-factor pattern
recognition model has been established. Furthermore, the prediction rules of outburst probability of coal and gas have
been determined. By adopting multi-factor pattern recognition probabilistic prediction methods, the regional forecast of
dangerous areas of coal and gas has been finished; risky areas, threatening areas and areas without obvious dangers of
coal and gas inside the coal field has been divided; assessment has been made on the outburst danger of coal and gas; the
accuracy of gas disaster prediction has been improved. The establishment of a relatively scientific prediction method of
outburst areas of coal and gas would make it possible to enable coal mine safety workers to make accurate judgment and
prevent the outburst of coal and gas.
Image encryption using spatial chaotic map
Fuyan Sun,
Zongwang Lü
Show abstract
In recent years, the chaos based cryptographic algorithms have suggested some new and efficient ways to develop secure
image encryption techniques. In this paper, we propose a new approach for image encryption based on spatial chaotic
map. The new scheme employs the spatial Cat map to shuffle the positions and uses another spatial chaotic map to
confuse the relationship between the cipher-image and the plain-image. The results of experimental, statistical analysis
and key space analysis show that the proposed image encryption scheme provides an efficient and secure way for realtime
image encryption and transmission.
Palm print image processing with PCNN
Show abstract
Pulse coupled neural networks (PCNN) is based on Eckhorn's model of cat visual cortex, and imitate mammals visual
processing, and palm print has been found as a personal biological feature for a long history. This inspired us with the
combination of them: a novel method for palm print processing is proposed, which includes pre-processing and feature
extraction of palm print image using PCNN; then the feature of palm print image is used for identifying. Our experiment
shows that a verification rate of 87.5% can be achieved at ideal condition. We also find that the verification rate
decreases duo to rotate or shift of palm.
Laser diffraction effect from liquid surface wave
Show abstract
Based on the laser technique,a method of detecting liquid surface wave was proposed.In this experiment, the intensity of
liquid surface wave was controlled by the electromagnetic exciter and the frequecy of the exciter was controlled by the
signal generator;When a He-Ne laser beam was obliquely incident upon the liquid surface wave directly, laser diffraction
patterns from liquid surface waves can be collected by charge-coupled device at the same time. The experiment
phenomena were visible and accordant to the theory well. The experiment results indicate that the proposed method has
the advantages of high sensitivity, high speed and non-destructive.
Automatic event detection of abandoned and stolen objects in real-time surveillance
Show abstract
This paper describes an approach aimed at automatic identification of events of abandoned and stolen objects in variety
environment. Our approach mainly include three steps of data processing: the first processing phrase is object extraction,
involving a dual-time background subtraction algorithm which dynamically updates two sets of background. Then,
extracted objects are classified as static or dynamic objects and human or non-human objects. Finally, a decision-making
model is employed to calculate a confidence score for the classification about event, and an alarm will be automatically
triggered if the certainty of action is higher than a pre-defined threshold. Also, the robustness and efficiency of the
method was tested on our real time video surveillance system and evaluated by public database such as AVSS 2007
datasets.
Image categorization based on spatial visual vocabulary model
Yuxin Wang,
Changqin He,
He Guo,
et al.
Show abstract
In this paper, we propose an approach to recognize scene categories by means of a novel method named spatial visual
vocabulary. Firstly, we hierarchically divide images into sub regions and construct the spatial visual vocabulary by
grouping the low-level features collected from every corresponding spatial sub region into a specified number of clusters
using k-means algorithm. To recognize the category of a scene, the visual vocabulary distributions of all spatial sub
regions are concatenated to form a global feature vector. The classification is obtained using LIBSVM, a support vector
machine classifier. Our goal is to find a universal framework which is applicable to various types of features, so two
kinds of features are used in the experiments: "V1-like" filters and PACT features. In almost all experimental cases, the
proposed model achieves superior results. Source codes are available by email.
Kernel orthogonal local fisher discrimination for rotor fault diagnosis
Show abstract
In order to better identify the fault of rotor system, one new method based on kernel orthogonal local fisher
discriminant (KOLFD) is proposed.Considering kernel mapping and iteration-orthgonal idea,training data with
supervision information was mapped to kernel space, computed local with-class scatter and between-class scatter,
constructed kernel fisher discriminant function. To ensure the minimum reconstruction error during deimensionality
reduction, algorithm joined the orthonormal constraints condition,found optimal basic projection vector by iterative
orthogonal approach.Then testing data was mapped by this vector and got new data's class information by neighbor
classifier,and eventually realize fault diagnosis.The experiment of rotor fault diagnosis shows, KOLFD algorithm has
better effect to other manifold learning algorithm.
Correct a particular fisheye lens distortion quickly using the coordinate map table
Ming Yang,
Cuicui Yang,
Tiantian Meng
Show abstract
Fisheye lens had the characteristics of shorter focus and large field,but fisheye images were distorted.Current fisheye
distortion correction algorithms had some problem,such as,taking point from the scene was approximate,fitting process
was complex,the computational cost was large and the real-time character could not be guaranteed.In this paper,
firstly,obtain the formula according to fisheye lens imaging principle.Secondly,get a correction curve from the special
points' relationship between the image captured by specific fisheye lens and calibration image. Finally,use the curve to
obtain the pixel map table between the fisheye image and the target image. Thus realize the fisheye image distortion's
correction. The results show that it is a real-time,fast,accurate fisheye image correction method.
Coherency estimation based on spectrum Gaussian-Hermite moments
Gengxiang Li,
Bo Yang,
Mo Dai
Show abstract
The process of identifying regions with similar texture and separating regions with different texture is an essential step
towards identifying surfaces and objects in the image. This paper describes a method of processing discontinuities of the
gray level image. Moments are widely used in image analysis and pattern recognition. The Gaussian-Hermite moments,
as one kind of orthogonal moments, are proposed to estimate coherency of the image in this paper. The local image is
firstly converted into frequency domain from spatial domain, and then coherency is measured by matrix constructed of
Gaussian-Hermite moments of energy within frequency domain. Compared to other methods, such as coherency
estimation based on geometric moments, cross correlation, eigenstructure, semblance, or gradient vector field, the
experimental results show a good performance in feature representation, regions recognition and regions.
Construction of panoramic image mosaics based on affine transform and graph cut
Show abstract
Image-based rendering has been a popular technique to simulate a visually rich telepresence and virtual reality
experience. The construction of panoramic image mosaics is an indispensable step in image-based rendering systems like
QuickTime VR and Surround Video. The conventional methods for creating panoramic image mosaics with regular
photographic or video images use geometrical feature points and optimization to the overlapped areas of the two
consecutive images, and then align and mosaic the corresponding areas using the blending or stitching algorithm. This
paper introduces a novel and efficient method to build panoramic image mosaics. The proposed method divides the
overlapped areas of the consecutive images into several sub-areas. The feature point, whose gradient value of intensity is
the maximum in the sub-area can be found easily. After selecting these feature points, we warp the images using an
affine transformation based on point set matching. Then the graph cut algorithm is used to build the seamless image
mosaic which makes the overlapped areas containing no visible ghosting or blurred details. It is shown by the
experiments that the new method can obtain mosaics of high quality and reduce the computing time.
A novel diamond search algorithm for fast block motion estimation
Jin-xiao Yang,
Yong-bo Zhang,
Li-hao Huang
Show abstract
Video coding is a complex process, comprising a combination of spatial, temporal and statistical data reduction
techniques. Of these techniques, motion estimation taking advantage of inter-frame information redundancy plays the
most important role. So finding the most efficient motion estimation algorithm remains an important research problem.
In this paper, a New Diamond Search (NDS) algorithm is proposed to improve the efficiency of video coding. The NDS
algorithm applies Cross Search Pattern (CSP) in the very beginning search steps and switch using Large Diamond Search
Pattern (LDSP) and Small Diamond Search Pattern (SDSP), so as to avoid the problem of local optimum. Experimental
results show that the proposed NDS algorithm is obviously improved in search speed and search accuracy compared with
the Diamond Search algorithm. The NDS algorithm can achieve very close quality compared to full search but with
18.51times speedup. In term of speed, compared with the DS algorithm, the NDS algorithm can achieve more than
125% speedup.
Improving example-based super-resolution via clustering training sets
Show abstract
More image patches in a training set making it more time-consuming has become a holdback of the real-time application
of example-based super-resolution. The paper proposes a method which clusters these training set to accelerate the
procedure. Before the super-resolution, a clustering method is used to partition the middle-frequency components in the
training set into some subsets. During super-resolution, the distances between each matching patch of low-resolution
image and each subset of training set are computed. The subset with the minimum distance is selected to carry out farther
matching. This procedure goes along until a most matching patch is found. The high-frequency patch within the training
set relevant to the found matching patch is selected as the researching output, which is used for super-resolution of
objective image. Two examples are use to illustrate the performance of the proposed algorithm, one using a factitious
image obtained by blurring and down-sampling an original image, and another using directly a true image. The results
show the proposed method can reduce effectively the computational complexity.
Research of ROI image compression based on visual attention model
Show abstract
Region of interest (ROI) coding is important in applications where certain parts of an image are of a
higher importance than the rest of the image. Human vision system actively seeks interesting regions in images to reduce
the search export in tasks, such as object detection and recognition. Similarly, prominent actions in video sequences are
more likely to attract human's first sight than their surrounding neighbors. Based on the mechanism of HVS, this paper
proposes a model of the focus of attention for detecting the attended regions in video sequences. It uses the similarity
between the adjacent frames, establishes the gray histogram, selects the maximum similarity as predicable model, and gets
position of the focus of attention in the next fame. And on the application of an algorithm for visual attention the paper
shows the region of interest (ROI) coding in JPEG 2000. JPEG 2000 ROI coding is used in combination with an
algorithm for VA to provide a progressive bit-stream where the regions highlighted by the VA algorithm are coded as an
ROI and presented first in the bit-stream. It can be seen that there is an improvement in image quality centered on the
ROI although this is achieved at the expense of reduced quality in the background of the image.
GIS, GPS, RS, & Wireless and Optical Communications in Industrial Engineering
Study of urban space buffer's applications on GIS
He Bing,
Peng Ye
Show abstract
In this paper, the buffer zone research methods analysis and define the path of time, the concept of travel
time and mathematical models.On the basis of the study and analysis of the Maoming City's urban space at different
stages with the tool of GIS,we summaries the law of the space of Maoming City's development.
Rotation-invariant texture retrieval by combining log-polar and nonsubsampled contourlet features
Xianqiang Zhu,
Zhenfeng Shao
Show abstract
This paper analyzes the spectrum influence between Radon transform and Log-polar transform when rotation effect is
eliminated. The average retrieval performance of wavelet and NSCT with different retrieval parameters is also studied.
Based on which, the authors design a multi-scale and multi-orientation texture transform spectrum, as well as rotationinvariant
feature vector and its measurement criteria. Then a new two-level rotation-invariant texture retrieval algorithm
based on no-parameter statistic features is proposed. Experiments on Brodatz image database show that the algorithm
proposed in this paper is appropriate for main orientation capturing and detail information description. Besides, the
combination of this two-level progressive retrieval strategy and multi-scale analysis method can effectively improve
retrieval efficiency compared with traditional algorithms and ensure a high precision as well.
Analysis of four schemes of polarization mode dispersion compensation in a 40Gbit/s optical communication system
Show abstract
With the development of optical fiber communication, and after chromatic dispersion and attenuation are compensated
for, polarization mode dispersion (PMD) in optical fibers becomes more and more important and cannot be ignored when
the transmission system is up to 40Gbit/s or above, or transmission distance is several thousand kilometers long. In the
paper, it is discussed four kinds of compensation schemes in 40Gbit/s optical communication system, it point out that
transmission distance is different when use different compensation schemes by pulse broadening theory. And when the
transmission speed is very high or the transmission distance is very long, it is necessary to compensated high-order PMD.
It compared 2nd-order PMD compensated system and first-order PMD compensated system, it point out that
2nd-order PMD compensated system is much better than first-order PMD compensated system by pulse broadening
theory.
Application of MAPGIS in extracting structure lines from magnetic data
Haijun Liu,
Yongping Yi,
Guochuang Hu,
et al.
Show abstract
Magnetic data are an important way to help experts to define the boundaries of magnetic substances. A traditional
method to extract structure information from magnetic data is to withdraw structure lines by hand. It was always a
troublesome and tedious work and it was a waste of time to modify when necessary. Therefore it is necessary to find an
efficient way to do this work. The development of computer and GIS makes it possible to process magnetic data
automatically. In this paper, we firstly analyzed the physical significance of magnetic first horizontal derivative as well
as the physical significance second vertical derivative. Secondly we elaborated the relationship between the structure
lines and the extreme lines. Then we explained in detail how to extract extreme lines from the first horizontal derivative
and how to filter the extreme lines in GIS. As an example we applied MAPGS6.7 to withdraw extreme lines from the
first derivative contour diagrams of a certain place. We adopted plus-minus correlating, direction correlating and
derivative correlating to filter the extreme lines and finally we mapped out the inferred structure. The ways explained in
this paper make it more effective and more convenient for experts to process magnetic data. The result in this paper
provides powerful reference for forecasting favorable place for prospecting.
An efficient and provably secure proxy signature scheme
Show abstract
Proxy signature is a special signature, it allows an original signer to delegate her signing capability to a proxy signer and
the proxy signer can produce a signature on behalf of the original signer. At present, most of proxy signature in essence
consists of two signatures. To overcome the problem, we propose a short efficient proxy signature scheme based on a
certificateless signature scheme. And we show that the proposed scheme is secure in the random oracle model. The
security of the scheme is related to Inverse Computational Diffie-Hellman Problem and the k-CCA problem. Comparison
with Huang et.al scheme, our scheme has an advantage over Huang et.al's scheme in terms of the size of proxy signature.
Since the length of proxy signature in our scheme is 160bit, it is very suitable for mobile devices.
Evaluation and study on PMD performance of OPGW optical cables in wind induced vibration and galloping test environment
Jie Li,
ZiYuan Zhao
Show abstract
This paper describes the concepts of wind induced vibration and galloping of power transmission lines, indicates the
necessity of evaluation in wind induced vibration and galloping test environment, proposes the evaluation method of
wind induced vibration and galloping, summarizes and analyzes the measured PMD performance data of OPGW optical
fiber obtained from wind induced vibration and galloping tests on OPGW, and reaches a conclusion that the PMD
performance parameters of OPGW cable stocks meet the industry standard in wind induced vibration and galloping
environment, and will play an important role in project construction guidance and operation maintenance.
Simulation analysis of modulated lidar on optical carrier for target detection in deep ocean
Show abstract
In the deep-ocean target detection, under water contrast is limited by incoherent backscatter clutter, and unable to meet
the requirements. Modulated lidar on optical carrier technology which detects the target in deep-ocean can effectively
inhibit the seawater backscatter and improve target's contrast. The system of modulated lidar on optical carrier for target
detection in deep-ocean is simulated by computer, and the results show that detection effects are significant improved,
and seawater parameters and modulation parameters have significant impact on the detection performance.
Measuring land subsidence by PALSAR interferometry in Yanzhou coal mine area
Show abstract
In this paper, we used PALSAR data to measure the land subsidence in Yanzhou coal mining area based on InSAR
technique. Firstly it simply analyzed InSAR technique and introduced some orbital radar satellites for InSAR measuring.
It introduced the PALSAR data for test and the selection of InSAR pairs. Then it illustrated the procedure of 2-pass DInSAR
data processing taking true data for example. It can be divided into inteferometric processing, simulating the
interferogram according to DEM, differential processing and filtering, phase unwrapping, conversion of deformation.
Then it got the deformation map of coal mining subsidence. It interpreted and analyzed the D-InSAR results in details. It
generated the isoline maps of land subsidence in coal mining area. We found that the maximum land subsidences in
vertical direction in two areas are up to 51.8 and 71.4 cm in 46 days. The test proves that PALSAR and SRTM DEM
data are suitable for monitoring the land subsidence in areas with vegetated land cover, such as coal mining area.
Detection of precipitable water vapor during summer rainfall by Nanjing GPS network
Haohan Wei,
Jiazhu Zheng,
Xiaoyun Shi
Show abstract
According to the ground-based principles of GPS meteorology, utilizing on the observation data of Nanjing regional
GPS network, introducing the observation data of three IGS base stations to construct very long baselines, and
combining with the atmospheric parameters by surface observation, the GPS absolute precipitable water vapor (PWV)
and relative PWV were estimated. Compared with SONDE PWV and the actual surface rainfall, results show that there
has a good relativity between GPS PWV and SONDE PWV; according to SONDE PWV, whether the average
difference or correlation coefficient, GPS absolute PWV is better than relative PWV; The variation of GPS PWV value
can be an intuitive response to changes in the actual surface rainfall, and Rainfall usually occurs in 4-6 hours after the
value of GPS-PWV had a rapid increase; and with the value of GPS PWV significantly rise or drop, it shows the
beginning or end of the rainfall. In addition, it's useful for studying on regional changes in atmospheric water vapor
with Estimating PWV by regional GPS network, and can improve the accuracy of meso-, micro-scale weather forecast.
Comparative experiments of road extraction from Google Earth imagery, QuickBird imagery, and IKONOS imagery
Show abstract
Main roads were detected using object-oriented method from Google Earth imagery, compared with QuickBird image of
the same region and the same capture time and IKONOS image of the same region. All images extract roughly the same
region of main road, and also similar image segmentation parameters and establishment of the classification rule sets.
Experimental results show that using Google Earth images to extract road could meet applications of less precision
demanding.
Research on the scattering of normal incidence surface of cone
Show abstract
The cone is divided into facets and each of the facets was analyzed with Lambert scattering distribution. The scattering
of rotating ellipsoid was detailed analyzed and the scattering model of rotating ellipsoid deduced at normal incidence. At
the same time, the applied model of the scattering was set up and the simulation was carried out with different contour
parameters at normal incidence, and the laser scattering curves were plotted. The results show that the scattering light
intensity achieve the maximum value when the zenith angle equals to 0 degree and decreases as zenith angle increasing,
when the zenith angle equals to 180 degree the scattering light intensity close to zero. The further analysis shows that the
scattering light intensity distributes smoothly when the ratio of long and short half axis is small but distributes more and
more acute as the ratio of long and short half axis gets greater.
New filtering method for GPS/INS integrated navigation
Xiao-chun Zhong,
Tao Song
Show abstract
The error caused by linearization in the filtering process is an important factor that decrease the filtering accuracy, and if
the error exceeds the allowable range it may result in filter divergence for some highly nonlinear systems. In the
GPS/INS system integrated with position and velocity, though the linearization process is eliminated because of its linear
model, the navigation accuracy is degraded by the correlation between the system's measurements. So the integration of
range was adopted for observation model, which makes use of the original measurements of GPS, so it can improve the
navigation accuracy. As this observation model is nonlinear, for the high accuracy requirement of GPS/INS integrated
navigation system, the UPF (Unscented Particle Filter) algorithm was introduced to estimate the system states. The UPF
is a nonlinear filer which is more suitable for nonlinear systems. The simulation was carried out under the condition of
less than four GPS satellites data available, the results of which show that the estimated values with UPF are quite close
to real values, and the UPF algorithm has a very good performance when it applied in the GPS/INS integrated
navigation.
An improved distributed routing algorithm for Benes based optical NoC
Show abstract
Integrated optical interconnect is believed to be one of the main technologies to replace electrical wires. Optical
Network-on-Chip (ONoC) has attracted more attentions nowadays. Benes topology is a good choice for ONoC for its
rearrangeable non-blocking character, multistage feature and easy scalability. Routing algorithm plays an important role
in determining the performance of ONoC. But traditional routing algorithms for Benes network are not suitable for
ONoC communication, we developed a new distributed routing algorithm for Benes ONoC in this paper. Our algorithm
selected the routing path dynamically according to network condition and enables more path choices for the message
traveling in the network. We used OPNET to evaluate the performance of our routing algorithm and also compared it
with a well-known bit-controlled routing algorithm. ETE delay and throughput were showed under different packet
length and network sizes. Simulation results show that our routing algorithm can provide better performance for ONoC.
Comparative study on 3D optical network-on-chip architectures
Jun Zhang,
Huaxi Gu
Show abstract
Network-on-chip (NoC) has been proposed to overcome the bottlenecks that on-chip communication faced due to the
increase of chip complexity. However, limitations such as long delay and high power consumption make traditional NoC
using electronic interconnects be inefficient for future use. A new design concept based on optical interconnects, Optical
NoC, which can provide low latency, high bandwidth and lower power dissipation, is a promising solution for future on-chip
interconnection. In this paper, we will investigate three different 3D Mesh-based ONoC topologies and compare
them in terms of End-to-end delay (ETE delay) and network throughput under varying injection rates. The simulation
result shows that 3D Partially-connected Optical-crossbar Mesh can achieve smaller network latency and throughput
than other three Mesh based architectures.
Analysis of the characteristic of optical communication system influenced by first-order polarization mode dispersion
Show abstract
When attenuation and chromatic dispersion are compensated for, polarization mode dispersion (PMD) is a major
limitation to the transmission capabilities of optical communication systems. So it is necessary to study PMD. In the
paper, we simulated first-order PMD statistical properties, the relationship of Mean DGD and transmission distance, the
deterioration of the receiver pulse broadening induced by PMD on pulse broadening theory, the distance relay L limited
by system transmission speed B under different PMD, several systems of different PMD restrict the transmission
distance etc., come to conclusion that PMD have some restrictions on high-speed system design. And the higher the rate,
the shorter the distance relay. For low-speed system, by choosing low values of fiber PMD can increase distance relay,
but for high-speed system, it is not a good idea to improve the system by selecting the optical fiber. In the actual design,
you should consider PMD index of optical fiber, chosen code pattern of the system, transmission rate and error rate etc.
A combination strategy for tracking the serial criminal
Show abstract
We build a Geographic Profiling Model to generate the criminal's geographical profile, by combining two
complementary strategies: the Spatial Distribution Strategy and the Probability Distance Strategy. In the first strategy, we
designate the mean of all the known crime sites as the anchor point, and build a Standard Deviational Ellipse Model,
considering the effect of landscape. In the second strategy, we take many factors such as the buffer zone and distance
decay theory into consideration and calculate the probability of the offender's residence in a certain area by using the
Bayesian Theorem and the Rossmo Algorithm. Then, we combine the result of two strategies and get three search areas
suit different conditions of the police to track the serial criminal. Apply the model to the English serial killer Peter
Sutcliffe's case, the calculation result shows that the model can effectively be used to track serial criminal.
A study of the environmental information acquisition system based on smart phones
Lingyan Xu,
Feixiang Chen,
Shaoliang Ni,
et al.
Show abstract
In this paper, it proposed a new environmental information acquisition system based on smart phones (Smartphone /
Pocket PC) which combined with Geographic Information System (GIS), Global Positioning System (GPS), wireless
communication technology in allusion to the current actual situation of environmental protection information acquisition
in city environmental protection department. System architecture and working principle is analyzed, and it designs the
main modules of the software and hardware. In addition, transport protocols and application of the implementation
method have been discussed. Experiments show that the environmental information acquisition system has high
precision, easy to use, information transfer with high efficiency and reliability. Not only have that, the paper also
discusses the effective strategies of network transmission of data encryption and the image transmission rate
improvement. In brief, it can effectively enhance the work efficiency of the city environmental protection department
when they collect relevant information.
The rough expression for topological relation of GIS entity based on binary relation
Show abstract
The relation for one entity and another entity of GIS is topological relation, and the set of two GIS entities with
topological relation is an ordered pair, and so it is a binary relation. We use some characteristic of binary relation and
rough set to study the relations for GIS entities, and we discover that the front, behind, upper and lower neighborhood
relation of GIS entity have even more extensive sets than itself based on binary relation's generalized approximation
space. We also discover that the upper and lower approximation of GIS entity for binary relation is more extensive than
Pawlark rough set. We discover that the upper approximation for a single-point set of one GIS entity is equal to it's front
neighborhood and is equal to it's behind neighborhood relation's inversion too.
Studying the space evolution of city block with GIS
Guang-lu Tu,
He Bing
Show abstract
There were many methods of urban design, this paper tried to explore a new method. First, I definited the notion of the
space evolution of city block, then took Dongchangkou as an example, on one hand, I studied the space evolution of this
block from three aspacts, such as construction form, system and ideology, on the other hand, I used GIS to analyse the
role and future direction of this block, at last, through the study, I gave a reasonable planning and design to
Dongchangkou. So, a new urban design what based on the study of space evolution of city block with GIS is put
forward, and its impact to the future urban design is predicted.
Comparison of Aster and SPOT4 modified soil adjusted vegetation index data over northwestern China
Show abstract
Vegetation indices (VIs) are essential parameters widely used in the biosphere remote sensing retrieval, and the
relationship between the same vegetation indexes derived from different sensors is critical to long-term monitoring of
land surface properties. In this paper, MSAVI data derived from visible and near-infrared data acquired by the ASTER
and SPOT4 sensors were compared over the same time periods and pixel size. The results showed that the two VIs play a
higher correlation in high data field. ASTER MSAVI is more sensitive to vegetation coverage information. SPOT
MSAVI overvalues the local vegetation reflection signals significantly. The linear relationship between vegetation
coverage and MSAVI requires field sampling data to complete correction.
Study on optical attenuation performance of special stock power optical cable based on a wind induced vibration environment in laboratory
Jie Li,
ZiYuan Zhao
Show abstract
For the purpose of 10G communication system upgrade for Guangdong Power Grid, laboratory simulation tests on
dynamic and temperature cycle are performed for the reserved cables (stock optical cables) of existing 2.5G special
optical cable lines that have operated for ten years, in order to verify the possibility of optical cable to be upgraded to a
10G transmission level and evaluate the degradation level of optical cables. This paper points out the necessity of
laboratory test on attenuation performance in a wind-induced vibration environment, describes the test methods thereof,
summarizes and analyzes a variety of optical attenuation performance data, and finds that the attenuation performance of
current OPGW, ADSS, ADL optical fiber lines in wind-induced vibration environment meets the industry standards.
Industrial Decision Support and Simulation Systems, Sensor Technology, and Intelligent Monitoring and Control/ICT Applications
Research on extension element model of robot
Junying Tian,
Honglin Chen
Show abstract
After having analyzed robot systems, this paper presents a method of organizing robot system with the matter-element in
extension theory. The robot physical structure determines the matter-element model of robots. The objective external
environment where robot is placed is plentiful, but the information that the robot can get about objective external
environment is limited. The sensors carried by the robot is the only source of information. The sensor data can describe a
robot external environment. The move, how to move, movement direction and movement mode of the actuators on the
information obtained by the sensors closely relate to external environment. If we ignore the sensors and actuators in a
robot system, the robot system is a processor and controller with moving targets and motion control rules. The moving
targets is an active control system that is independent. In an extension element model of robots, the sensor data describe
the inputs of robots, the movement of actuators describes the outputs of robots, and the robot itself is a controller with
moving targets and motion control rules. The study on the matter-element models of robots can develop new research
methods and ideas of robots, which is important for the robot's development and application.
Applied research on extension element model in hydraulic system
Junying Tian,
Bingyang Wei,
Jianhai Han
Show abstract
This paper further researches the application of extension element model of hydraulic system. A extension element
model of hydraulic system contains three parts: the prototype of a hydraulic component, the matter-element and relationelement
model of hydraulic components, and the matter-element and relation-element model of hydraulic systems. The
foundation building a extension element model of hydraulic system is to build its database of the prototypes of a
hydraulic component. The matter-element and relation-element model of hydraulic components is a model for physical
hydraulic components. The the matter-element and relation-element models of hydraulic components on a hydraulic
system can describe the internal links among hydraulic components. Through analysis of the connections among the
matter-element and relation-element models of hydraulic components, rings, power bond graphs, and dynamic equations
will be found and gained. Thus, the performance and digital simulation for a hydraulic system will be realized and
finished.
Application of improved grey prediction in intelligent valve positioner
Show abstract
Intelligent valve positioner is the main part of pneumatic actuator. Intelligent valve positioner exhibits nonlinear and
time-varying properties that cannot be captured by any existing model, so it is difficult to satisfy its requirement for high
accuracy and fast speed through the use of a single control method. We propose a fuzzy-PID control method based on a
grey model, which combines fuzzy control with PID control, and the system provides inherently more accurate
predictions. To address the potential problems of the fast increasing serie's dimension and an over-capacity of micro
controller unit (MCU) as a result of frequent measurements by the grey model, we introduce a grey model of equal
dimension and the recursive least-squared method, which keep the matrix dimension constant, reduce the computation
load and ensure real-time results. Our simulations confirm that our method provides rapid response, high accuracy and
superior anti-interference capability. Our experiment proves that the intelligent valve positioner can be successfully
applied to solve real-world problems.
Centralized robust fusion estimation in estimation of paper basis weight based on norm-bounded parameter uncertain model
Xue-bo Jin,
Jia Bao,
Hai-jiang Zheng
Show abstract
This paper gives the models of uncertain multisensor system based on norm-bounded parameter uncertain model method,
by which, the centralized robust fusion estimation is developed for uncertain multisensor system based on Linear Matrix
Inequality (LMI) methods. The developed fusion estimation methods are used for paper basis weight estimation
considering the uncertainty of paper machine in practice. The results of experiments show the application of information
fusion theory can develops the estimation performance, especially, the performance of state estimation can keep well
even when the sensor fails.
A new method for simplification of discernibility function
Dao-lin Wang,
Guan-ming Peng
Show abstract
Discernibility matrix and discernibility function are mainly used for attribute reduction on information system .
Simplification of the discernibility function are converted to the same solution transform of linear logical equation
group. Coefficient matrix of linear logical equation group can be transformed into the simplest matrix with the
elementary row transformation. Then a new method which divides the attribute into absolutely necessary attribute,
relatively necessary attribute and absolutely superfluous attribute is put forward, and a heuristic algorithm based on
attribute classification is proposed.
The research of distributed interactive simulation based on HLA in coal mine industry inherent safety
Zhi-wu Dou
Show abstract
To solve the inherent safety problem puzzling the coal mining industry, analyzing the characteristic and the application
of distributed interactive simulation based on high level architecture (DIS/HLA), a new method is proposed for
developing coal mining industry inherent safety distributed interactive simulation adopting HLA technology.
Researching the function and structure of the system, a simple coal mining industry inherent safety is modeled with
HLA, the FOM and SOM are developed, and the math models are suggested. The results of the instance research show
that HLA plays an important role in developing distributed interactive simulation of complicated distributed system and
the method is valid to solve the problem puzzling coal mining industry. To the coal mining industry, the conclusions
show that the simulation system with HLA plays an important role to identify the source of hazard, to make the measure
for accident, and to improve the level of management.
Fuzzy neural network model applied in the mine water inrush prediction
Jian-yu Xiao,
Min-ming Tong,
Qi Fan,
et al.
Show abstract
Combining with fuzzy reasoning, neural network and back propagation algorithm, a fuzzy neural network model for the
prediction of mine water inrush is built after the analysis of the main factors affected the mine water inrush, such as
water pressure, fault throw, water conducted zone width, aquifer thickness and confining bed thickness. Then, on the
basis of the water inrush mechanism and some practical examples, influential factors on mine water inrush are divided
into three groups, and after that 81 fuzzy inference rules are constructed effectively. And then, the four-layer back
propagation fuzzy neural network takes effect in training the input variables. Finally, a simulated prediction of the fuzzy
neural network is made by using test samples. The results of simulation show that mine water inrush model based on
fuzzy neural network is superior to the traditional BP neural network on training speed and predicting precision.
CT discrimination and image process on damage process of unsaturated compacted loess during triaxial creep
Xiaojun Li,
Lihua Jiang,
Yichuan Tang
Show abstract
The triaxial creep compression tests of compacted loess samples are conducted with a new set of modified triaxial
compression apparatus. With the new apparatus, the loess sample can be scanned with CT machine at the same time
during compression process. The different damage process of compacted loess sample is directly observed for the first
time with CT images and CT numbers. The initiation mechanisms of loess micro-crack during different creep
compression processes are analyzed with CT images.
A mathematical model for dynamic over-load protection based on the heat balance differential equation
Changchun Chi
Show abstract
The study of new digital algorithm of over-load protection is meaningful to improve the efficiency of the over-load
protection validity. This paper establishes a mathematical model for the dynamic over-load protection, based on the heat
balance differential equation, to determine how temperature increases with the change of the protection object's current.
Different from the common inverse time characteristic, this model conforms to the load actual operating law and the
protection request.Wavelet analysis, as a new kind of time-frequency representation technique, has been applied widely
in different engineering practice. To solve the problem of the full-wave Fourier algorithm which is disable to filter
decaying DC component and has worse frequency characteristic, it is presented an algorithm depending on the
combination of subtraction filter and full-wave Morlet complex wavelet amplitude algorithm. It can filter efficiently
decaying DC component, has clear frequency characteristic, and make calculation with high precision.
Study on workshop layout of a motorcycle company based on systematic layout planning (SLP)
Show abstract
The method of SLP has been applied in a motorcycle company's layout planning. In this layout design, the related
graphics have been used to illuminate the logistics and non-logistics relationships of every workshop to get the integrated
relationships of workshops and preliminary plans. Comparing the two preliminary plans including logistics efficiency,
space utilization, management conveniences, etc, an improvement solution is proposed. Through the improvement
solution, the productivity has been increased by 18% and the production capacity is able to make 1600 engines each day.
Classifying human activities using feature points
Hao Zhang,
Zhijing Liu,
Qing Wei,
et al.
Show abstract
This paper presents a new classification method for single person's motion, which is represented by Haar wavelet
transform and classified by Hidden Markov Models. What it solves is that the feature points are detected by Haar
wavelet transform. We extract binary silhouette and segment them by cycle after creating the background model. Then
the low-level features are detected by Haar wavelet transform and principal vectors are determined by Principal
Component Analysis. We utilize Hidden Markov Models to train and classify cycle sequences, and demonstrate the
usability. Compared with others, our approach is simple and effective in feature point detection, as the advantages of
Haar wavelet transform detector lying in computational complexity. So the video surveillance based on these is
practicable in (but not limited to) many scenarios where the background is known.
String vibration measurement system based on PSD
Show abstract
Equipment monitoring, fault diagnosis, and noisiness analysis and eliminating and so on industrial production processes
can't do without vibration measurement, this paper compose a vibration test system with the laser - string - PSD sensor
and the corresponding signal processing electric circuit. PSD is one optics semiconductor sensor, can output the different
size electric current according to the luminous spot in its different superficial position, fast locate with precision to the
luminous spot position, carry on experiment to PSD sensor in different vibration source and frequency, obtained
vibration waveform of the luminous spot on PSD photosurface, calculates each kind of vibration parameter about
vibration source, such as vibration frequency and amplitude, experimental result is in accordance with the actual
parameter, verifies PSD not only to have own good qualities in the position measurement, but also have the unique
superiority in the vibration measurement.
Extension research on robot external environment
Junying Tian,
Jianhai Han,
Bingyang Wei
Show abstract
After having analyzed robot systems, this paper puts forward a matter-element model of robot external environment.
This matter-element model is built by sensor data in robot system. The element establishing matter-element model is
basic element, which is defined by a branch matter and is regarded as a special matter-element model. The operation +
not only produces a new matter-element model between two basic elements, but also produces a new matter-element
model between two matter-element models. A system, which is made up of the set of matter-element models and the
operation +, constitutes a semigroup. This semigroup is a mathematical model of robot external environment on extenics.
The elements in the semigroup can describe various forms of robot external environment, and the relation among these
forms, and various shapes of robot system.
Gear faults diagnosis based on wavelet-AR model and PCA
Zhixiong Li,
Xinping Yan,
Chengqing Yuan,
et al.
Show abstract
Gear mechanisms are an important element in a variety of industrial applications and about 80% of the breakdowns of
the transmission machinery are caused by the gear failure. Efficient incipient faults detection and accurate faults
diagnosis are therefore critical to machinery normal operation. The use of mechanical vibration signals for fault
diagnosis is significant and effective due to advances in the progress of digital signal processing techniques. Through
virtual prototype simulation analysis and experimental study, a novel method for gear multi-faults diagnosis was
presented in this paper based on the wavelet-Autoregressive (AR) model and Principal Component Analysis (PCA)
method. The virtual prototype simulation and the experimental test were firstly carried out and the comparison results
prove that the traditional Fast Fourier Transform Algorithm (FFT) analysis is not appropriate for the gear fault detection
and identification. Then the wavelet-AR model was applied to extract the feature sets of the gear fault vibration data. In
this procedure, the wavelet transform was used to decompose and de-noise the original signal to obtain fault signals, and
the fault type information was extracted by the AR parameters. In order to eliminate the redundant fault features, the
PCA was furthermore adopted to fuse the AR parameters into one characteristic to enhance the fault defection and
identification. The experimental results indicate that the proposed method based on the wavelet-AR model and PCA is
feasible and reliable in the gear multi-faults signal diagnosis, and the isolation of different gear conditions, including
normal, single crack, single wear, compound fault of wear and spalling etc., has been effectively accomplished.