This PDF file contains the front matter associated with SPIE Proceedings Volume 11144, including the Title Page, Copyright information, Table of Contents, Author and Conference Committee lists.

For the measurement of three-dimensional (3D) shapes, active optical measurement systems based on pattern projection are widely used. These sensors work without contact and are non-destructive. Between one camera and the projector or between two cameras, the 3D reconstruction is performed by detection and triangulation of corresponding image points. Recently, we developed a 3D stereo sensor working in the visible range of light (VIS). It consists of two highspeed cameras and a GOBO projection-based high-speed pattern projector. Our system allows us to successfully reconstruct 3D point clouds of fast processes such as athletes in motion or even crash tests. Simultaneously measuring the surface temperature would be of great benefit as fast processes usually exhibit local temperature changes. In order to include thermal data into the evaluation, we have extended our existing high-speed 3D sensor by including an additional high-speed long-wave infrared (LWIR) camera. The thermal camera detects radiation in the spectral range between 7.5 and 12 μm. We map the measured temperatures as texture onto the reconstructed 3D points. In this contribution, we present the design of this novel 5D (three spatial coordinates, temperature, and time) sensor. The simultaneous calibration process of the VIS cameras and the LWIR camera in a common coordinate system is described. First promising measurements of an inflating airbag, a basketball player, and the crushing of a metal tube performed at a frame rate of 1 kHz are shown. Keywords: high-speed visible (VIS) and infrared (IR) cameras, 3D shape measurement, multimodal sensor, GOBO projection, aperiodic sinusoidal patterns, temperature mapping

Through the acquisition and processing of several spectral channels within the multispectral data, the demands on signal processing and data handling increase enormously. With the help of intelligent signal pre-processing on programmable system on chip platforms (pSoC), captured data can be corrected and evaluated directly after image acquisition. PSoC combine the advantages of freely programmable logic (FPGA) and sequential processor systems (ARM technology) and significantly increase the integration density of embedded image processing systems. However, the design effort for these systems is increasing strongly, so that hardware/software co-design approaches must be used for implementation. The paper covers the design methodology, the implementation and the evaluation of multichannel acquisition systems using multispectral image sensors as an example.

A novel sensitivity-enhanced intrinsic fiber Fabry-Perot interferometer (IFFPI) high temperature sensor based on a hollow- core photonic crystal fiber (HC-PCF) and modified Vernier effect is proposed and experimentally demonstrated. The all fiber IFFPIs are easily constructed by splicing one end of the HC-PCF to a leading single mode fiber (SMF) and applying an arc at the other end of the HC-PCF to form a pure silica tip. The modified Vernier effect is formed by three beams of lights reflected from the SMF-PCF splicing joint, and the two air/glass interfaces on the ends of the collapsed HC-PCF tip, respectively. Vernier effect was first applied to high temperature sensing up to 1200°C, in this work, and the experimental results exhibit good stability and repeatability. The temperature sensitivity, measured from the spectrum envelope, is 14 to 57 times higher than that of other configurations using similar HC-PCFs without the Vernier effect. The proposed sensor has the advantages of high sensitivity, good stability, compactness, ease of fabrication, and has potential application in practical high-temperature measurements.

The proposed laboratory exercise is focused on the measurement of indoor environment quality (temperature, humidity, CO2 concentration) using digital measurement system based on Modbus protocol. Students acquire practical knowledge and general principles of the Modbus protocol, configuration and communication principles of smart sensors and measurement of temperature, humidity, CO2 concentration together with power consumption. The benefit for students in comparison with the old task is the improved quality of the laboratory task in the area of measurement circuits and fieldbus communication with the focus on industrial application.

Depth sensors for three-dimensional object acquisition are widely used and available in many different sizes and weight classes. The measuring method used and the measuring accuracy depend on the task to be performed. The integration of depth sensors in mobile devices such as tablets and smartphones is largely new. The TrueDepth system of the iPhone X shows which measurement accuracies can be achieved with these systems and which areas of application can be achieved in addition to consumer fun. The investigations show that the TrueDepth system of the iPhone X can be used for measuring tasks with accuracies in the millimeter range.

Supercontinuum "White Light Lasers" have become a well-established turn-key fiber-laser technology addressing a wide range of applications from bio-medical-imaging to optical device characterization, where they add value due to their unique combination of optical parameters, including an extremely wide spectral coverage from 400 nm to 2400 nm, several W of optical output power, and a perfect Gaussian beam that can be focused down to the diffraction limit. This review introduces the state-of-the-art technology and some applications.

Hyperspectral imaging as a method for in-line product assessment has evolved into an industrial technique widely used for spectral characterization of materials. The currently available computing performance allows the execution of data processing and chemometric analysis in milliseconds and enables the use of hyperspectral imaging in real time. Potential applications and the corresponding requirements are manifold. Separation of different materials, detection of defects and quantification in quality management are a few examples of major areas of operation. The broad range of potential applications is illustrated by examples covering different fields such as recycling, food or construction materials and different wavelength ranges of the used hyperspectral cameras.

With the appearance of cameras containing newly developed polarized image sensors, which include an on-chip polarizer array placement, there is an upcoming need to characterize these cameras according to a uniform standard. These polarizer arrays consist of 2×2 matrices with four different angled polarizers, placed on top of each pixel. Therefore, this setup differs from the traditionally used multi camera and multi-polarizer setups. In terms of characterizing cameras, the EMVA 1288 standard already provides a widely established guideline on how to characterize monochrome and color digital cameras with linear photo response characteristics. The aim of this paper is to characterize a polarization camera, which takes advantage of these newly developed sensors, within the framework of the EMVA 1288 standard. Therefore, a measuring setup is constructed to be able to isolate the response characteristics of the individual pixels of the polarizer arrays. The measuring setup is built in accordance to the EMVA 1288 standard, thus consisting of a monochromatic light source, a measurement tube and a photodiode. In addition, a rotatable linear polarizing filter is being placed in front of the image sensor, to be able to control the polarization direction of the incident light. The results obtained will then be evaluated and interpreted, and proposals for an extension of the EMVA 1288 standard for the characterization of polarization cameras will be given, based on the experience gained throughout the characterization.

Tissue survival depends on the hemoglobin in blood to deliver oxygen to support cellular respiration. Without oxygenated blood surrounding tissue dies. Oxygenated hemoglobin has a distinct visible – near infrared absorption spectrum compared to deoxygenated hemoglobin. A hand-held multispectral reflectance imaging device, SnapshotNIR, was designed to provide a measure of the relative attenuation of reflected light from oxygenated and deoxygenated hemoglobin. The device combines the use of near real-time image classification / segmentation and regression to effectively determine the ratio of oxygenated to deoxygenated hemoglobin in the superficial vascular bed being imaged. This measurement of hemoglobin oxygen saturation can help detect local deficits in oxygen delivery to tissue. This simple, hand-held, battery-powered imaging device can quickly survey large areas of tissue in a complete non-invasive fashion. It is easy to use and suitable for use in the operating room or clinics that have limited infrastructure. Examples are given on the performance of the device in general surgery and in the assessment of peripheral circulation.

The aim of this work is to develop possibilities of an effective audit procedure shall be worked out. Audits are systematic, independent and documented processes for obtaining evidence. This ensures conformity with the requirements. In order to achieve certification according to standards, an external system audit must be carried out by an accredited company. These are very time-consuming and costly because the auditor has to be on site. A concept of how the audit can be carried out without an on-site appointment is developed and evaluated. The possibilities of current technologies - such as VR - are also considered and integrated into the concept. Due to the high-level structure this can be carried out on an integrated audit which combines different management systems.

With the advent of industry 4.0, the introduction of smart manufacturing and integrated production systems, the interest in 3D image-based supervision methods is growing. The aim of this work is to develop a scalable multi-camera-system suitable for the acquisition of a dense point cloud representing the interior volume of a production machine for general supervision tasks as well as for navigation purposes without a priori information regarding the composition of processing stations. Therefore, multiple low-cost industrial cameras are mounted on the machine housing observing the interior volume. In order to obtain a dense point cloud, this paper reviews aspects of metric stereo calibration and 3D reconstruction with attention being focused on target-based calibration methods and block matching algorithms.

Two activities, how to increase interest in the study of electrical engineering and electronics, are described in the paper: workshops for teachers and participation in the Embedded Technology Club for students. Workshops for teachers acquaint them with the utilization of IT in a laboratory education, to make it more attractive and so to increase the interest of students to study natural sciences or engineering. The educational low-cost laboratory system and its applications in laboratories are presented during the workshop. The Embedded Technology Club (ETC) is a club intended as for students of secondary schools, who has an interest in electronics, as for other persons interested in this area. The business is focused both on education in the fundamentals of up-to-date design, development, and programming of electronic devices and solving some more simple projects concerning instrumentation. Activities are designed so that also students of secondary schools, who have no experience with electronics, but only basic knowledge in physics, can take part in. The priority is to have the interest to learn new things and to get hands-on experience.

Automatic Optical Inspection (AOI) is an important part of modern electronics manufacturing. While profilometric technologies are very present in inspection systems specialized on surface mounted technologies (SMT), through hole technology (THT) oriented inspection still mostly relies on two-dimensional data. Solder joints of THT connections have highly reflective metallic surfaces and thus are a very challenging object for typical light based profilometric measurement systems. A new approach is presented, which uses specular reflections of angled illuminations to reconstruct the surface of THT solder joints for inline inspection. The observed reflections depend on the position of the angled light sources, as well as the surface angle of the inspected object and its relative position to the observing camera. The reflections of each angled illumination are observed separately and processed as a stack of images. To reduce the complexity the image stack is transformed into a polar coordinate system leveraging the rotation symmetry of THT solder connections. The surface angle of every observed pixel is estimated based on the observation stack and corrected for its position relative to the camera and illumination. The surface angle then gets integrated along the transformed coordinate axis. This results in the estimation of the surface profile of the solder connection. After retransformation into the source coordinate system, this can be used to classify the solder joint for defects and insufficient solder volume.

Bearings are precision machine elements with very high requirements regarding the roundness, waviness and roughness of the rolling elements as well as the bearing rings. The production processes have to meet sub-micron tolerances. To verify production quality, state of the art tactile measuring technologies are used, such as roundness and waviness measurements as well as roughness measurement according to ISO 4287. To measure roundness and waviness a special gauge is used, where as a second gauge is measuring the roughness of the surface. The measurements are mostly performed in a laboratory environment checking parts in a random sampling procedure. To meet the rising demands on quality of bearings, it is necessary to increase the sample size. On the other hand, this increases the time needed for quality inspection. The optical measuring technology SCATTERING LIGHT can help to solve these contradictory demands. The technology is based on the reflection of light from a surface. It can be used to measure roundness, waviness and roughness in one single operation. Providing a clean environment, the measurements with scattering light can be automated and integrated in production processes such as honing and grinding. Furthermore, the cycle time is shorter than from the tactile measuring procedure. Results from scattering light measurements can be calibrated. Roundness and waviness are traceable to international standards. The optical result for surface roughness “Aq” is a new parameter that does not correlate to the common known Ra and Rz values but to the occasionally used Rdq value.

In many coating-lines exact thickness-control is essential. We demonstrate some results with a stroboscopic imaging ellipsometer, which enables to acquire an image of the thickness distribution of a certain layer during or right after production making it an ideal tool for inline-monitoring. The prototype obtains a thickness-distribution with 10-100Hz acquisition rate. The accuracy is in the range 2-10% of the layer thickness. Since there are no moving parts and only monochromatic illumination with a standard laser-diode the solution is cost-effective and can be easily installed. We demonstrate three different applications, namely SiO₂ on Si, oil on steel, and MgF₂ on glass.

In this paper we report the development and test of a high sensitivity fiber Bragg grating (FBG) seismic sensor system for intrusion detection application. A high sensitivity FBG seismic sensor is designed and its sensitivity is above 1000pm/g. Unbalanced Michelson interferometer and Phase generated carrier(PGC) algorithm are adopted by the demodulation system. The system noise is below 10^-3 pm/ √Hz and the minimum detectable seismic signal is below 1μg√Hz . The FBG seismic sensor has a maximum detection range of about 70 meters for human and about 200 meters for small wheeled vehicle.

The term deep learning is almost on everyone's lips these days, in the area of computer vision manly because of the great advances deep learning approaches have made amongst others in object detection and classification. For general object location or classification tasks there do exist several giant databases containing several millions of labeled images and several thousands of different labels like COCO and ImageNet. In contrast in industrial applications like quality inspection there do hardly ever exist such training data not only for reasons of confidentiality of trade secrets. An obvious way to remedy this deficiency is the synthetic creation of image data. Physically based rendering attempts to achieve photorealistic images by accurately simulating the ow of light of the real world according to various physical laws. Therefor multiple techniques like Ray Tracing and Path Tracing have been implemented and are becoming increasingly widespread as hardware performance increases. The intent of this article is to give a wide but nevertheless preferably comprehensive overview which approaches have been pursued in recent literature to generate realistic synthetic training images. The development of various rendering methods from rasterization to bidirectional Monte Carlo path tracing is outlined, as well as their differences and use. Along with the terminology a few mathematical foundations like the Bidirectional Reflectance Distribution Function (BRDF) are briefly described. Altogether specially concern is given to industrial data and quality control, comparing literature and the practical application of its results.

Platform F0-Lab including LEO (Little Embedded Oscilloscope) was designed and developed to substitute a professional set of instruments in experiments concerning electronics, measurement, and instrumentation. It is based on a microcontroller, which includes the sufficiently fast ADC. It enables to realize (in connection with PC) a simple DC voltmeter, a digitizing oscilloscope, a waveform recorder, and a PWM signal generator. The microcontroller can be placed in a solderless breadboard where the student realizes experimental circuits. The PC controls this virtual instrument, displaying results and supplying the power of the entire set using USB. F0-Lab is successfully used for three years and more than 300 various implementations have been already realized.

This paper presents a principle for scene-related camera calibration in Manhattan worlds. The proposed estimation of extrinsic camera parameters from vanishing points represents a useful alternative to the traditional target-based calibration methods, especially in large urban or industrial environments. We analyse the effects of errors in the calculation of camera poses and derive general restrictions for the use of our approach. In addition, we present methods for calculating the position and orientation of several cameras to a world coordinate system and discuss the effect of imprecise or incorrectly calculated vanishing points. Our approach was evaluated with real images of a prototype for human-robot collaboration installed at ZBS e.V. The results were compared with a perspective n-Point (PnP) method.

Feature detection in pointclouds is becoming increasingly important for a wide range of applications. For example fields of robotics, autonomous driving and medical image processing are using acquisition sensors with 3 dimensional output. These points are usually defined by x, y, and z coordinates and represents the outer surface or shell of an object. Challenges are to develop methods for effective ways to get just the relevant information in pointclouds and accelerate postprocessing steps. The approaches presented in this paper are based on innervolumetric 3D data, which are described by chaotically organized points. The data used are generated by the layer-by-layer acquisition and composition of individual point clouds in additive manufacturing. The data is merged, preprocessed and subsequently the characteristics are extracted.

Publisher’s Note: This paper, originally published on 17 September 2019, was withdrawn per author request.

This paper presents the design and simulation of a single-shot optical 3D sensor based on multispectral pattern projection and a stereo-vision setup of two multispectral snapshot cameras. The performances of various combinations of available multispectral cameras, spatial light patterns, and 3D reconstruction algorithms as well as the geometric arrangements of the stereo-vision camera setup are simulated and analyzed. This simulation-based investigation delivers two optimized combinations of sensor components in terms of hardware and algorithm as orientations for practical sensor development in future, and an appropriate arrangement of the stereo-vision setup is determined. Moreover, the influences of sensor noise on 3D reconstruction are also estimated.

This work presents, summarizes and validates a fast, accurate and general measurement and modeling technique to obtain spectral near field data of LED systems in order to improve the optical design process of modern high quality LED systems. It requires only a minimum of goniophotometric near field measurements as well as no time-consuming angularly resolved spectral measurements. The procedure is named physically motivated basis spectra (PMBS) as its main assumption is that each piece of angularly and spatially varying spectral information can be described as the weighted sum of its physical basis spectra such as the individual semiconductors or a phosphor. Based on detailed spectral information regarding the goniophotometric measurement setup, the spectral model is obtained by solving a simple system of linear equations using the obtained near field measurements. The complete process is validated and applied to different state-of-the-art LED systems. The obtained results can be used directly in state-of-the-art ray tracers.

Passive-stereo-imaging as a part of three-dimensional image processing can be applied for many use cases where a depth information is necessary. This is often the case when classical two dimensional image processing get to its borders. Passive-stereo needs at minimum two images of the same object of interest. An initial one-time system calibration is absolutely essential. The usage of more than two sensors offer more system stability and reduce hidden image elements. A variant of semi-global-matching algorithm is used to find the matching points in the image pair and is implemented in the FPGA. The lateral shift of two matching points is called disparity. All disparity values together result in a disparity image which is the basis for a point cloud. By using embedded technologies, a compact and portable acquisition system could be realized. Programmable system-on-chip (PSoC) combining FPGA and ARM computing power in one chip design. This technology is an ideal solution to so acquire the images from the image sensor and calculate the disparity images.

Artificial intelligence and machine learning are becoming increasingly important in science and society. In image processing, they are mainly used for object classification. The aim of this paper is the comparison of classical supervised machine learning methods with innovative deep learning (DL) approaches in terms of performance, which is described by the calculated accuracy. Classifiers of different characteristics are used. These are the Support Vector Machines, Random Forest, k-Nearest-Neighbor, and Naive Bayes. They are compared to two not pre-trained and four pre-trained neural networks. The former neural network are based on LeNet, the second ones include AlexNet, GoogleNet and ResNet provided by Matlab as well as a pre-trained neural network provided by MVTec HALCON. Comparisons were made using the recognition rates achieved with five real data sets from industrial applications. The results showed that not pre-trained neural networks produce worse results than classical classifiers with the given small amounts of data for training. On the other hand, the pre-trained networks achieved surpassing recognition rates. However, if there are features that describe the classes very well, the recognition performance of classical machine learning methods little differs from that of deep learning algorithms.

Filter-On-Chip CMOS sensor equipped cameras are a convenient, reliable and affordable approach for the parallel acquisition of spatial and spectral information. The combination of pixel-arranged spectral filter matrices on CMOS sensors increases their integration density and system complexity by several times compared to standard RGB cameras. Due to their system design, these cameras have an increased spectral crosstalk and specific dependencies from the angle of radiation. The paper will show how to develop and set up a measurement arrangement for the characterization of the channel specific spectral sensitivities under different angles of radiation. These characterizations are necessary to develop a more robust model for the camera pixel-value correction to ensure the comparability and reproducibility of the measured values. Therefore, a measurement setup to investigate the influence of the angle of incident radiation on filter-on-chip CMOS sensors was developed. After initial investigations with a setup in which the camera was simply rotated and investigations with a lens and changed f-number confirmed that the angle influence results in a measurable difference in the sensor response, a new measurement arrangement was developed to investigate this behavior more precisely. The developed measurement arrangement allows multispectral resolving image sensors to be radiated with collimated light at reproducible angles of incidence and with adjustable wavelengths. By comparing the measured values with illuminances measured using a calibrated photodiode in the same setup and with the same parameters, it is possible to evaluate the angle dependence based on quantum efficiency curves according to the EMVA 1288 standard. The investigations carried out, the developed principles and the realized semi-automatic measurement arrangement will be shown and explained to characterize the capabilities of multispectral resolving filter-on-chip CMOS sensor equipped cameras for applications in industry and biomedicine.

This paper presents an algorithm detecting automatically a wide variety of cracks on monochrome images of concrete floors. It is part of a vision system, generating a crack map to support the condition monitoring for buildings. The suggested method uses successively radiometric, geometric and contextual information. An automatic supervised adaptive intensity-threshold method handles radiometry information. A threshold-based method was chosen because it can separate even thin cracks from the background. In order to cope with different image cluster intensities, first a clustering algorithm separates regions of different intensities, and second the threshold is adaptive at the cluster level; being a continuous function of the cluster intensity. At some points, function values were learned via supervised classification. Then, we performed an interpolation between these points in order to get a threshold whatever the cluster intensity (continuity). At this step, we have a thresholded image. However, due to texture, non-crack dark defects and dirtiness we have false positives. In order to overcome this problem, connected pixels are grouped into regions, and after discarding small regions, a size-dependent geometrical shape filter is suggested. The shape of a crack region depends on its area, and this relation was empirically learned. A region is retained if its shape features are above some area depending thresholds. However, mainly due to texture, we have still false positives because some non-crack entities have radiometry and geometry of cracks. Fortunately, they are often small isolated regions and are discarded via an isolation filter. Tests performed on many images show very encouraging results.

Electromobility has been a strongly growing market for years. This is the reason for the demand for battery technology and electric components continue to increase. In these technologies, the material copper is indispensable due to its high electrical conductivity. With the same electrical conductivity, substituting aluminum to copper leads to a reduction in costs and weight. For example, substitution is not possible at connecting points, therefore dissimilar aluminum-copper joints are highly important. In micro processing, pulsed laser beam welding is applied to achieve a slender weld seam. However, the mixing of both joining partners leads to the formation of intermetallic phases during welding. This requires a precise detection of the process stage in order to limit the weld seam depth close to the interface between both materials. In this paper, a pulsed laser welding process between aluminum and copper was tested by using a fiber laser (IPG-YLM- 450/4500-QCW, pulse duration < 10 ms). The optical spectrum of the welding process was detected by spectrometers in the visible light range. When aluminum is welded with copper, the wavelength spectrum changes due to the material dependent emission. The maxima within the wavelengths of each joining partner could thus be determined and transferred to photodiodes with suitable bandpass filters. This leads to an increase of the temporal resolution during the measurement compared to spectrometers, allowing the analyzation of the time-related signal characteristics. A difference between heat conduction welding and deep welding as well as the transition from upper to lower sheet metal could be determined.

In this work, we present a calibration procedure of a multispectral snapshot camera, which is based on a multi-aperture system approach combined with a slanted linear variable spectral filter. The ultra-compact multispectral imaging system exploits state of the art micro-optical manufacturing techniques on wafer-level, which leads to a size of only 60 × 60 ×28 mm³. The setup enables the single-shot acquisition of 66 spectral channels with a linear spectral sampling over an extended wavelength range of 450-850 nm and a spatial sampling of 400×400 pixels per channel at a large field of view of 68°. In particular, we propose a spectral and spatial calibration procedure in order to extract hyperspectral data cubes from the acquired raw image and further, to analyze characteristic system parameters. Finally, we demonstrate the systems capabilities for advanced object classification using characteristic spectral indices by utilizing a customized multispectral analysis.

In this paper, we present a novel fiber optic ultrasonic sensing system to conduct a 2D temperature field monitoring. The fiber optic ultrasonic sensing system was used as an ultrasonic pyrometer to measure the temperature field. The ultrasonic pyrometer was based on the thermal dependence of the speed of sound in air. The speed of a sound wave traveling in a medium was proportional to the medium’s temperature. A fiber optic ultrasonic generator and a microphone were used as the ultrasonic signal generator and receiver, respectively. A carbon blackPolydimethylsiloxane (PDMS) material was utilized as the photoacoustic material for the fiber optic ultrasonic generator. A test was performed outside of a lab furnace, the testing area temperature range was from 26°C to 70°C. A 2D temperature field was mapped. The 2D temperature field map matched with the reference thermocouple results. This system could lead to the development of a new generation temperature sensor for temperature field monitoring in coalfired boilers or exhaust gas temperature monitoring for turbine engines.

Encoding information using magnetic microstructures is utilised e.g. in magnetic stripe cards or banknotes. In order to analyse these structures quickly and non-destructively a magneto-optical setup based on the Faraday effect can be used. The Faraday effect states that the plane of polarisation of polarised light is rotated when it passes a Faraday-active material and is subjected to an external magnetic field. Measuring the change of intensity of light passing a set of polarisers allows the calculation of the change of the polarisation angle, which in turn allows drawing conclusions on the magnitude of the external magnetic field. A first setup yielded very good results in the high-speed analysis of patterns with a structure size of about 50 μm. However, the setup's low amplitude resolution only allowed qualitative measurements. In order to find the limitations of this measurement principle with respect to amplitude, temporal, and spatial resolution as well as their interdependence a new setup was purpose-built for characterisation. Its components were examined closely and various methods of signal enhancement were evaluated. The measurements displayed long- and short-term temporal as well as spatial dependencies. The subsequent enhancement of the signal's amplitude resolution came with a loss in temporal or spatial resolution and vice- versa. The interdependence of amplitude, spatial, and amplitude resolution was characterised further and from this a generalised description of their lower bound for a given set of parameters was derived. This should serve as an estimate of the feasibility as well as a build guideline of a similar setup.

Multi-spectral camera set-ups may generally allow for creating surveillance applications even under unfavorable conditions, such as low-light environments or scenes involving vastly different lighting conditions. A high- resolution color camera, a high-dynamic-range camera and an infrared thermal camera were combined into a self-sufficient platform for continuous outdoor operation. The sheer amount of produced data poses a serious challenge, both in terms of available bandwidth and processing power, because self-sufficiency requires using relatively low-power components, and privacy, as high-resolution, multi-spectral image data are sensitive information. Thus, relevant objects of interest had to be efficiently extracted, tracked and georeferenced on the sensor platform. These data, from one or more sensorheads, are then sent via WLAN or mobile data link to a central control unit, possibly anonymized, e.g. prompting immediate action by a human operator in a disaster response use case, or stored for further offline analysis when used in the framework of "Smart City". Applying the classic stereo vision approach would require calibrating both intrinsic and extrinsic parameters of all cameras. The input data's multi-spectral nature complicates the correspondence problem for extrinsic parameter calibration and subsequent stereo matching, while intrinsic parameter calibration according to the pinhole camera model is made difficult due to the cameras having to be focused at infinity. However, by making certain reasonable assumptions about the observed scene in typical use cases, accepting a possible loss in localization accuracy, camera calibration could be limited to the bare minimum and less computational power was required at run-time.

Concerning hyperspectral imaging, there is a need for handling big hyperspectral data. Especially merging data that was captured by different sources is very complex. Due to the different properties in resolution and other technological aspects, there are requirements for calibration, preprocessing and merging functions. This work is a summary of the approach that was developed in the framework of the Qualimess research project for handlings these demands. The invented declarative programming model is an innovative solution for implementing algorithms for all these aspects in an easy to handle signal flow model. The algorithm is designed in a way that it uses as little computing resources as possible, which is of high importance when processing and analyzing spectral cubes with standard computing systems. In this last development stage, previous investigations regarding the fusion of hyperspectral data captured by push-broom imaging systems are extended to other imaging technologies and hyperspectral 3D-imaging as well. Furthermore, there is a presentation of the developed graphical user interfaces that can be used for analyzing the data from an application-oriented point of view. These have the aim of extracting as much information as possible and presenting it to the user in a comfortable way. Finally, results in the framework of applied measurement under use of this signal processing method are presented for the first time. By evaluating this, an assessment can now be made of it under which circumstances these algorithms can be used in the context of industrial applications.

Uncertainty evaluation plays a key role in assessing and comparing measurement results, e.g. towards the design of experiments and the design of measurement systems. Nevertheless, it is frequently neglected by engineers particularly in early design phases, which may cause problems during the design process or even lead to system failures. In prior work, the use of software tools for uncertainty calculation in measurement science education has been suggested in order to raise awareness and increase the use of such tools to obtain GUM conform uncertainty estimates. However, we found that uncertainty of complex-valued quantities represents a major challenge in this context. Therefore, we present a proposal to extend the teaching concept towards such quantities, based directly on the utilization of an "Uncertainty Toolbox" for MATLAB maintained by our research group. The approach evaluates the uncertainty in complex parameters considering the real and imaginary components separately, with potential correlations between them arising from shared input quantities. As teaching example we study the Maxwell-Wien Bridge, as it is commonly taught in measurement science courses and brings in aspects such as above mention correlations and different representations of the measurement result (magnitude/phase, real/complex part of impedance). Based on this example, advantages and disadvantages of the presented teaching philosophy are discussed, emphasizing how problems arising from uncertainty may be identified in early design phases also considering complex-valued quantities.

In traditional confocal microscopy, there is a trade-off between spatial resolution and field of view due to the limitations of objectives. To solve this problem, diffractive optical elements (DOEs) with overlapping apertures are used to generate high-NA illumination spots in a large area. However, currently such DOEs can only be used as illuminators which are not suitable for 3D surface measurements. In this work, the idea of superposition is utilized to expand the scope of application of the DOEs. These DOEs are designed by simulation and tested in the experiments. The results show that the proposed DOEs can be used in 3D surface measurements and have the potential to solve the problem of high-NA objectives.

LumiScan is a novel technology by which highly accurate measurements on shiny, metallic parts are feasible without any special requirements on illumination. The system is fully operational on the shop floor, demonstrating its superior performance.

The one-week intensive preparatory course “Practical Electronics” focused on measurement and instrumentation, which is designed especially for school-leavers of secondary schools (high schools in the USA), is described in the article. In many cases students, who are coming at the CTU in Prague, Faculty of Electrical Engineering, have only a vague idea of electronics despite they use electronic devices every day. The course is organized before the 1^st semester and only basic secondary education in physics is supposed. It is divided between a theoretical part and a laboratory work. Each day starts with two hours lecture, where the actual laboratory work is explained. The following activities run in laboratories. A set of necessary components, solderless breadboard, connecting wires, and a low-cost multimeter are prepared for each student. While the 1^st day is devoted to the connection of the simplest electric circuits and voltage measurements using a low-cost multimeter, the 2^nd day starts with connecting of the virtual measuring instrument (VI) based on a microcontroller, which includes a simple DC voltmeter, a digitizing oscilloscope, and a pulse generator. It is used in tasks solved in the following part of the course.

Currently, Deep Learning (DL) shows us powerful capabilities for image processing. But it cannot output the exact photometric process parameters and shows non-interpretable results. Considering such limitations, this paper presents a robot vision system based on Convolutional Neural Networks (CNN) and Monte Carlo algorithms. As an example to discuss about how to apply DL in industry. In the approach, CNN is used for preprocessing and offline tasks. Then the 6- DoF object position are estimated using a particle filter approach. Experiments will show that our approach is efficient and accurate. In future it could show potential solutions for human-machine collaboration systems.

The subject of the Industrial Electronics and Sensors Laboratory has been designed to be focused on the practical part of the teaching. Here we describe its concept, used equipment and task examples. Experience gathered during subject first run is reported, including final students’ feedback.

Lens undistortion and image rectification is a commonly used pre-processing, e.g. for active or passive stereo vision to reduce the complexity of the search for matching points. The undistortion and rectification is implemented in a field programmable gate array (FPGA). The algorithm is performed pixel by pixel. The challenges of the implementation are the synchronisation of the data streams and the limited memory bandwidth. Due to the memory constraints, the algorithm utilises a pre-computed lossy compression of the rectification maps by a ratio of eight. The compressed maps occupy less space by ignoring the pixel indexes, sub-sampling both maps, and reducing repeated information in a row by forming differences to adjacent pixels. Undistorted and rectified images are calculated once without and once with the compressed transformation map. The deviation between the different computed images is minimal and negligible. The functionality of the hardware module, the decompression algorithm and the processing pipeline are described. The algorithm is validated on a Xilinx Zynq-7020 SoC. The stereo setup has a baseline with 46 mm and non-converged optical axis between the cameras. The cameras are configured at 1.3 Mpix @ 60 fps and distortion correction and rectification is performed in real time during image capture. With a camera resolution of 1280 pixels × 960 pixels and a maximum vertical shift of ± 20 pixels, the efficient hardware implementation utilizes 12 % of available block RAM resources.

As the most important characteristic gas, carbon monoxide (CO) can be used for early detection of coal spontaneous combustion in mine goafs. Conventional gas analysis system for coal mine combustion monitoring is chromatography- based gas tubing bundles system, which suffers from long time delay. In this report, a sensitive and stable CO monitoring system was developed by using a distributed feedback (DFB) laser operating at 2.33 μm and a Herriott-type multi-pass gas cell with a 20-m optical length, taking advantage of the in-situ monitoring, excellent accuracy and simple structure available from direct absorption spectroscopy. The detection accuracy of system was about ±0. 2 ppm when as low as 1 ppm CO gas was detected. And data monitored can be used to determine that the detection limit of system was about 0.2 ppm. Further, a long-term continuous monitoring evaluation has clearly demonstrated the long-term stability and reliability of the monitoring system. The results obtained have validated the potential use of such a CO monitoring system in a practical monitoring application, such as the coal spontaneous combustion monitoring.

Built on a design developed from an advanced mathematical model, a practical fiber optic sensor, which is an analog of the familiar ‘hot-wire’ wind velocity monitor is developed, as an intrinsically-safe sensor device for coal mining monitoring applications. The underpinning optical fiber-based principle used is the shift in the center wavelength of a Fiber Bragg Grating which is cooled by the gas flowing over it and the device sensitivity found was determined to be ~1370pm per unit m/s wind velocity (in the range of 0-0.57 m/s), ~109pm per unit m/s in the range 0.57-2.26 m/s and ~33pm per unit m/s in the range of 2.26-5.66 m/s. In this paper, the factors that influence the device response time, such as the sensor probe surface heat transfer coefficient, wind (gas) velocity and pump power have been investigated in the laboratory. It was found that the greater the surface dissipation factor of the sensor, the shorter the response time, furthermore, the response time was observed to decrease as the wind velocity increased. A method of further shortening sensor response time using wind speed variation slope is proposed.

This work describes the detection of multiple spectra on a single matrix-sensor. The source of the spectra is a newly developed miniaturized monolithic spectrometer. It can detect up to twelve separate channels. The spectra are located next to each other on the matrix-sensor. Each spectrum contains of multiple spots in a discontinuous line. Due to the small size strong distortions of the spectra are expected. Additionally, the spectra can overlap partly. This leads to a challenging process of matching each of the single spots to the right spectrum.

White light interferometry is a major optical non-contact and therefore nondestructive testing method for nanostructures and surface reconstruction. By the reason of scanning, the data throughput is high and the resulting data stack can exceed gigabytes of raw data. Effective data compression was realized in an FPGA early in the signaling cascade. On the one hand this can significantly boost the achievable data throughput from the sensor, on the other hand the compression results in fragmented raw data with non-equidistant sampling steps and is therefore incompatible with FFT based reconstruction algorithms. In order to face this issue the fragmented l1- norm transform (flot) was developed. The flot reconstruction algorithm is a symbiosis of the l1-norm known from compressive sensing and additionally the wavelet-transform. In contrast to the traditional wavelet-transform the flot algorithm has no dependence on FFT and can quickly handle non-equidistant sampled data. Raw data is heavily independent between pixels in white light interferometry by design. Therefore, implementing the re- construction algorithm on massively parallel hardware is promising. In the last decade this usually meant data processing on GPUs. Nowadays alternatives in the form of affordable CPU clusters or easy to program FPGAs gain importance. OpenCL is a framework to accelerate highly parallel problems on all of the three platforms. In this paper the implementation of the flot algorithm in OpenCL will be explained, compared by speed and power consumption and categorized for suitable use-cases.

A polarization state detector (PSD) measures the state of polarization of the detected light. The state of polarization is fully described by the Stokes vector containing four Stokes parameters. A division-of-amplitude photopolarimeter (DOAP) measures the four Stokes parameters by simultaneously acquiring four intensities using photodetectors. A key component of the DOAP is the first beam splitter, which splits up the incoming beam into two beams. The effect of the beam splitter on the state of polarization of the reflected (r) and transmitted (t) beam is determined by six parameters: R, T, ψ_r, ψ_t, Δ_r, and Δ_t. R and T are the reflectance and transmittance, and (ψ_r, Δ_r) and (ψ_t, Δ_t) are the ellipsometric parameters of the beam splitter in reflection and transmission, respectively. To measure the Stokes vector with high accuracy, the six optical parameters must be chosen appropriately. In previous work, the optimal parameters of the beam splitter have been determined as R = T = 1/2, cos² 2ψ_r = 1/3, ψ_t = π/2 - ψ_r, and Δ_r-Δ_t modulo π=π/2 by calculating the maximum of the absolute value of determinant of the instrument matrix. Using additional quarter-wave plates eliminates the constraint on the retardance and hence simplifies the manufacturing process of the beam splitter, especially when broadband application is intended. To compensate a suboptimal value of Δ_r-Δ_t, the azimuthal angles of the principal axes of the retarders must be adjusted, for which we provide analytic formulas. Hence, a DOAP with retarders is also optimal in the sense that the same values for the determinant and condition number of the instrument matrix are obtained. When using two additional retarders, it is necessary to install both on the same light path in order to obtain an optimal DOAP. We will show that is also possible to get an optimal DOAP with only one additional quarter-wave plate instead of two, if one of the Wollaston prisms is rotated.

Lenses have a number of faults. The distortion is an error which has a negative effect especially in metrological applications. In extreme cases, distortion can also be visible to the naked eye. The distortion of lenses is a systematic error that can be corrected numerically. For this purpose, a target with a smaller local deviation is necessary. These targets are produced and measured with great effort. Especially for the area of large object fields the effort increases enormously. For this reason, no correction has been made for larger measuring ranges. In order to be able to apply a correction nevertheless, a compromise solution is required that produces a target with little effort, with which a correction becomes possible that is better than not applying any correction at all. Here, calibration targets were printed and examined with commercially available printers.

New and revised standards are increasing the demands placed on management systems. The aim of this paper is to investigate how the effort required to create and maintain management systems can be minimized. A concept is presented, which also considers the integration of different management systems. First the existing management systems are evaluated. Then the requirements from practice are determined. These are examined and evaluated. Based on both results it is determined which requirements an integrated management system must fulfil. On this basis, a practical concept of an integrated management system is developed. As well optimal interfaces are considered. This concept is transferred into a web-based application. In order to ensure an optimal implementation and expandability, the structure is realized modularly.

Forestry in Germany and all over the world is undergoing a big change based on the digitalization. A combination of recent advances in measurement technology offers strong possibilities to use optical measurement systems as an addition to the conventional forestry machinery. Cameras, lasers, and sensors extend the accuracy of how we collect bio data, e.g. diameters, distances, heights, the total amount of wood volume, timber etc. Detailed data of the plant structure as well as calamities are the basis to get a better overview and improve the downstream process chain. Forestry is an economical field working with natural products, especially wood. Therefore, foresters need to determine the exact amount of the re-growing products. The certification systems like PEFC (Program for the Endorsement of Forest Certification Schemes) and FSC (Forest Stewardship Council) rely on precise data of wood volume, which has been harvested. Hence, nature product producer need more techniques and optical measurement systems, which are adapted to this very rough field of application. The cameras will be exposed to rain, fog, heat, cold, ice, mud, earth, dirt, stones, vibrations, and more effects in the outdoor use. This paper examines some of the recent projects and joint-ventures between forestry and photo-optical measurements and describes how they can be used to bridge the gap between two disciplines, addressing areas of both sides. We need to exchange our knowledge to find new ways of applying optical systems in Forestry.

Publisher’s Note: This paper, originally published on 17 September 2019, was replaced with a corrected/revised version on 17 February 2021. If you downloaded the original PDF but are unable to access the revision, please contact SPIE Digital Library Customer Service for assistance.