This PDF file contains the front matter associated with Volume 10780, including the Title Page, Copyright Information, Table of Contents, Authors and Conference Committee lists

The fast radiative transfer model RTTOV offers different ice cloud models to simulate observations from hyperspectral infrared instruments such as AIRS, IASI or CrIS. Since the last RTTOV version 12 there are two ice cloud models: the first one is an ice water content and temperature parameterization of a large dataset of ice cloud optical properties coming from the Met Office, and the second one is based on the ice crystal size dependent ice cloud optical properties database provided by SSEC. In order to compare these two ice cloud models, we used collocated data from the A-Train during two weeks of global observations. The ice cloud profiles description (ice water content and effective size of ice crystals) used as inputs for RTTOV is provided by the synergetic retrieval from active sensors (CloudSat and CALIOP/CALIPSO) named the 2C-ICE product. The RTTOV top of atmosphere simulations using these ice cloud profiles are then compared with collocated infrared observations from the IIR radiometer that is onboard CALISPO and has three infrared window channels. We found that RTTOV-12 is able to reproduce observations with nearly no biases and standard deviation below 7 K in window channels. The results show also the strong impact of the knowledge of the ice crystal size that is not currently information predicted by NWP models. By using hyperspectral infrared instruments, the spectral consistency of these two ice cloud models is also preliminary discussed.

Time-series of global satellite measurements allow for monitoring the global and regional environmental characteristics and associated change. Global surface skin temperature has been retrieved from MetOp-A/IASI hyperspectral infrared measurements over the past decade. Monthly and spatially-gridded surface skin temperature is produced to show some phenomena associated with its natural variability. The anomalies of surface skin temperature are used to estimate its recent decadal trend. Error estimation and evaluation has been performed and discussed in order to understand the uncertainty in the estimated trends. The trends of IASI global surface skin temperature anomalies are compared with those of the NASA Goddard Institute for Space Studies global surface air temperature anomalies. Despite the physical differences between surface skin and air temperature, reasonable agreement is shown between these two datasets indicating consistency and global surface warming during the past decade. The trend of IASI global surface skin temperature anomaly illustrates that an approximate 0.027°K/yr. global average increase has evolved during a decade long period of June 2007 – April 2018. This decadal warming trend is more pronounced in the northern hemisphere. This work demonstrates the utility of using Earth’s surface skin temperature derived from satellite measurements for monitoring global and regional surface characteristics.

Measurement system validation is critical for advanced satellite sensors to achieve their full potential of improving observations of the Earth’s atmosphere, clouds, and surface for enabling enhancements in weather prediction, climate monitoring capability, and environmental change detection. Field campaigns focusing on satellite under-flights with validation sensors aboard high-altitude aircraft provide an essential component important for performing such satellite measurement system validation. The NASA Langley Research Center National Airborne Sounder Testbed – Interferometer (NAST-I) is a cross-track scanning Fourier Transform Spectrometer system that is frequently deployed aboard NASA aircraft as part of the key payload sensors in validation and airborne science field experiments. One recent experiment, the Suomi NPP (SNPP) Arctic airborne field campaign (SNPP-2), was conducted out of Keflavik, Iceland between 7-31 March 2015 to address SNPP validation and JPSS risk mitigation for very cold scene observations and satellite sensor cross-validation (i.e. between the advanced satellite infrared sounders CrIS, AIRS and IASI) in the Arctic region. This paper addresses benefits achieved from such airborne validation field experiments and focuses on cold scene radiances observed during the SNPP-2 campaign; emphasis is placed on inter-comparisons between the NAST-I airborne observations and those from the Cross-track Infrared Sounder (CrIS) aboard the SNPP satellite and, with a particular focus on, handling the presence of non-uniform scene conditions.

Spaceborne infrared atmospheric sounders measure the spectrum of the upwelling radiance in the infrared with ultra-high spectral resolution. The resolution is sufficient to measure absorption features of atmospheric constituents enabling retrieval of atmospheric temperature and water vapor profiles, surface emission and atmospheric constituents. The Atmospheric Infrared Sounder (AIRS) on Aqua launched in May of 2002 was the first hyperspectral grating-based infrared sounder designed for this purpose and is still operational today. AIRS has been followed by the Infrared Atmospheric Sounding Interferometer (IASI) on MetOp A and B, and the Cross-track Infrared Sounder (CrIS) on Suomi NPP and JPSS. All instruments are operating well improving weather forecast and providing a wealth of information about the atmosphere. Additional CrIS and IASI instruments are expected to be launched providing data of this type into the late 2030’s. AIRS, CrIS and IASI are all Low Earth Orbit (LEO) instruments with nominal spatial resolutions of 14km. Future IR sounders must achieve higher spatial and temporal resolution to match improvements in forecast models and be less costly to match anticipated future budget pressures. Higher temporal resolution can be achieved in several ways including operation in Geostationary Earth Orbit (GEO) or in constellations of LEO satellites. Higher spatial resolution can be achieved using larger format focal plane assemblies in the instruments and larger aperture telescopes. Grating spectrometers are well suited to large format FPAs by allowing a wide field of view in a compact package. They also provide long life and are easy to operate. Concepts for next generation grating spectrometer IR sounders that have been developed over the years at JPL are presented along with technology advancements made to enable these concepts to achieve their stated goals.

The Advanced Technology Land Imaging Spectroradiometer (ATLIS) is a small (0.04 m3), multispectral pushbroom imager to provide visible through shortwave (VSWIR) calibrated imagery for the Sustainable Land Imaging-Technology (SLI-T) reference mission architecture (RMA) [1]. ATLIS is designed to provide imaging spectroradiometry that meets SLI-T RMA key parameters with an instrument that is much smaller and much less massive than previous land imaging systems. This paper describes a NASA ESTO funded project to design, build and test a six spectral band prototype ATLIS called ATLIS-P that will establish whether this compact, low mass design approach with wide field of view (WFOV), free form reflective telescope, large format, small detector digital FPA and on-chip processing meets SLI-T RMA VSWIR requirements.

Monitoring heavy metal stress on rice is of great significance for food security. In this paper, we used NDVI time series during the whole growing period of rice to identifying the rice growing differences under varied heavy metal stress. Here the NDVI time series were with high spatial-temporal resolution and obtained by blending MODIS and Landsat NDVI data. We extracted two kinds of features: Max NDVI value and time-integrated NDVI and use Fisher discrimination to explore the rice phonological differences under mild and severe stress levels. Results indicates that under severe stress the values of the metrics for presenting rice phonological differences in the experimental areas of heavy metal stress were smaller than the ones under mild stress. This means using the phenology differences can help to monitoring the heavy metal contamination.

Beibu Gulf is the main mangrove growth district in Guangxi province of China. Tieshan Harbor in the Beibu Gulf was chosen as the study area for this research. Based on the pixel spectral reflectance angle formed by the VIS/NIR bands in the multispectral remote sensing images of HJ-1A satellite, a red band angle vegetation index (RAVI) was proposed, which would be beneficial for separating vegetation on land and water. The mangrove judging standard was formed using a combination of RAVI and pixel band reflectance standard deviation (BStdev). Based on the indices analysis and growth area generation, the mangrove extraction mode was established, which effectively captured the mangrove spatial distribution from sparse to dense coverage and in different shapes along the coastline of Beibu Gulf.

Many studies have shown that hyperspectral measurements can help monitor crop health status, such as water stress, nutrition stress, pest stress, etc. However, applications of hyperspectral cameras or scanners are still very limited in precision agriculture. The resolution of satellite hyperspectral images is too low to provide the information in the desired scale. The resolution of either field spectrometer or aerial hyperspectral cameras is fairly high, but their cost is too high to be afforded by growers. In this study, we are interested in if the flow-cost hyperspectral scanner SCIO can serve as a crop monitoring tool to provide crop health information for decision support. In an onion test site, there were three irrigation levels and four types of soil amendment, randomly assigned to 36 plots with three replicates for each treatment combination. Each month, three onion plant samples were collected from the test site and fresh weight, dry weight, root length, shoot length etc. were measured for each plant. Meanwhile, three spectral measurements were made for each leaf of the sample plant using both a field spectrometer and a hyperspectral scanner. We applied dimension reduction methods to extract low-dimension features. Based on the data set of these features and their labels, several classifiers were built to infer the field treatment of onions. Tests on validation dataset (25 percent of the total measurements) showed that this low-cost hyperspectral scanner is a promising tool for crop water stress monitoring, though its performance is worse than the field spectrometer Apogee. The traditional field spectrometer yields the best accuracy as high as above 80%, whereas the best accuracy of SCIO is around 50%.

This study proposed a remote sensing-based approach to quantify the spatio-temporal patterns of vegetation dynamics and associated impact factors in typical karst regions of Southwest China. Google Earth engine (GEE), the world's most advanced geospatial data cloud computing platform, was employed to construct long time series satellite data set with 30 m resolution, composed of nearly 4,000 Landsat scenes from 1988 to 2016. Image preprocessing was also conducted on the GEE platform. The maximum value composite (MVC) method was used to produce annual maximum normalized difference vegetation index (NDVI) of the study areas. Annual maximum fractional vegetation cover (annFVC) was thus quantitatively estimated based on Dimidiate Pixel Model (DPM). Ordinary least squares (OLS) regression was adopted to identify the spatial patterns of the direction and rate of change in annFVC at a pixel scale. In addition, a terrain niche index (TNI) was used to investigate the influence of topographic factors on vegetation trends. Moreover, the relationships between annFVC and climatic factors were identified using correlation analysis. The results show that annFVC significantly increased at a rate of 0.0032/year in Nandong and 0.0041/year in Xiaojiang watershed for the period 1988-2016. Furthermore, 26.97% and 27.16% of pixels were found to undergo significant increase in terms of annFVC in Nandong and Xiaojiang, respectively. For both Nandong and Xiaojiang, decreasing vegetation trend was curbed with the increase of elevation and slope. Additionally, correlation analysis demonstrated that annFVC was more strongly and positively correlated with temperature than with precipitation in spite of insignificance.

Multiband LiDAR systems, which are typically single wavelength in transmission and reception, are becoming more applicable for scientific use. However, traditional LiDAR receivers do not scale well to tens or hundreds of received bands. We introduce the design for a spectrographic receiver using an array detector for laser spectrometers and present two of the many possible applications: fluorescence spectroscopy in the visible range and IR reflectance spectroscopy. Each laser pulse has the capability of exciting a target in various wavelengths, and a spectrographic receiver would be able to interpret this excitation, while a typical LiDAR consisting of single wavelength receiver would not. Using a spectrograph in a system with a pulsed laser in the visible or UV range is capable of the detection of fluorescent signal. These spectra reveal the presence of organics and is an applicable technology for planetary science. A spectrograph coupled with a pulsed laser in the IR range shows capability of detecting the presence of water in various forms also applicable technology for both Earth and planetary science. Both systems utilize a Czerny-Turner spectrograph design with a ZnSe prism for the dispersion of light onto an Avalanche Photo Diode (APD). This paper introduces the concept and design of a spectrographic receiver for laser spectrometers, as well as two possible applications.

Gases released from a volcano, such as sulfur dioxide (SO2) and carbon dioxide (CO2), present hazards to the environment and local populations as well as providing a means to monitor volcanic activity and study pre-eruptive signatures. In the Mid-Wave InfraRed (MWIR) from 3 to 5 microns both the aforementioned volcanic gases exhibit characteristic absorptions. Remote sensing in the MWIR, however, is challenging due to the limited amount of signal available to measure. This presents technical challenges on achieving high signal-to-noise ratios; therefore, acquiring adequate data in the MWIR has been difficult without cryogenically cooling the instrument. However, ecent improvements to microbolometer technology and emerging interferometric techniques have allowed us to acquire good thermal infrared data without the need for cooling. By utilizing the advantages of an imaging interferometer paired with an uncooled microbolometer, we demonstrate the use of a MWIR compact, hyperspectral imager for volcanic gas detection. The instrument, the Miniaturized Infrared Detector of Atmospheric Species (MIDAS), is representative of an instrument that could feasibly be flown on a small satellite in low earth orbit for the detection and monitoring of volcanic gases. Recently MIDAS was deployed to Kilauea’s Halema’uma’u pit crater which during the deployment had an active lava lake that was continuously releasing volcanic gases. Sources like the Kilauea lava lake provide high background temperatures that aid MWIR measurements of volcanic gases. We present hyperspectral analysis of volcanic gases from the Kilauea lava lake using data from the MIDAS instrument and line by line radiative transfer analysis. Brightness temperature maps of the lake surface show values consistent with direct thermocouple measurements and point radiometer measurements. Here we present resolved images of spectral radiance, brightness temperature, and CO₂ concentrations. The map of CO₂ is relatively uniform, but show subtle variation at the 2553 - 3313+/- 167 ppm level.

The intense development in imaging spectrometers and related technology has yielded systems that are highly performing. Current grating-based designs utilize focal plane arrays with aberrations controlled to a fraction of a detector element and low F-numbers for high étendue to maximize the signal to noise performance. Tailored grating facets using two or more blaze angles optimize the optical efficiency across the full 400-2500 nm solar reflective spectral range. Two commonly used forms, the Offner-Chrisp and Dyson designs, are adaptations of microlithographic projectors with a concave or convex mirror replaced by a shaped grating; maintain a high degree of spatial-spectral uniformity. These gratings are relatively difficult to manufacture using either e-beam lithography or diamond machining. The challenge for optical designers is to create optical forms with reduced size, weight, and power (SWaP) requirements while maintaining high performance. We have focused our work in this area and are developing a breadboard prototype imaging spectrometer that covers the full VNIR/SWIR spectral range at 10 nm spectral sampling, has a large swath of 1500 spatial samples, and is compact. The current prototype is for an F/3.3 system that is 7 cm long with an 8 cm diameter with aberration control better than 0.1 pixel assuming an 18 μm pixel pitch. The form utilizes a catadioptric lens and a flat dual-blaze immersion grating. The flat grating simplifies manufacturing and we are currently exploring the manufacture of the grating through grayscale optical lithography where the entire pattern can be exposed at once without stitching errors.

Minerals such as pyroxene, plagioclase, olivine, and ilmenite, which constitute most of the lunar surface rocks with varying size and shape, have distinctive spectral characteristics in the VNIR and SWIR regions. To analyze the composition of lunar surface minerals, several spectrometers based on AOTF was developed to detect lunar surface objects and to obtain their reflectance spectra and geometric images includes the Visible and Near-IR Imaging Spectrometer(VNIS) onboard China’s Chang'E 3 and Chang’E 4 lunar rover and Lunar Mineralogical Spectrometer(LMS) onboard Chang'E 5 and Chang'E 6 lunar lander. These spectrometers, which use acoustic-optic tunable filters as dispersive components, consist of a VIS/NIR imaging spectrometer, an SWIR spectrometer, and a calibration unit with dust-proofing functionality. They are capable of synchronously acquiring the full spectra of lunar surface objects and performing in-situ calibration. This paper introduces these instruments, including their working principle, implementation, operation, and major specifications, as well as the initial scientific achievement of lunar surface exploration.

While standard laser range finders use modulation signals, such as sharp pulses and periodic signals, to generate fast physical random numbers, our method does away with the modulator, and instead, utilizes laser diodes’ frequency noise and a frequency discriminator, to produce the intensity noise signals that generate fast physical random numbers. Observed through a frequency discriminator, beams having the same intensity noise patterns travel along two different paths, but with a time lag. We measured and calculated their cross-correlation, confirming the degree of difference in their optical paths, up to a distance of 50 m. We improved range resolution by taking advantage of the polynomial approximation of the coefficients around the peak of the correlation waveform.

A microscene is a hyperspectral image collected using a hyperspectral sensor mounted above a tray, typically in a laboratory setting. Materials can be placed in the tray and illumination controlled to either analyze the materials used or to simulate overhead (aerial or satellite) imagery. Choosing the materials allows simulation of overhead imagery in controlled experiments, for example mixtures and abundances of chemicals, materials as they undergo physical and chemical processes such as oxidation and weathering, and vegetation at different stages in environmental processes. Microscene imagery enables experiments in controlled circumstances not easily producible in overhead imagery. Moreover, the cost of collecting microscene imagery is a small fraction of overhead collection. Microscene imagery is an emerging technology, and in this paper we address an evaluation microscene imagery to determine how well it simulates overhead imagery, comparing microscene imagery of vegetation to overhead AVIRIS and HYDICE imagery over vegetation. We use statistical measures to compare microscene imagery to overhead imagery, including comparing material spectra, means, eigenvalues, the Mahalanobis distance between image means, and for the first time the Bhattacharyya distance between image covariances. The Bhattacharyya is a statistical measure of the distance between two statistical distributions, related to the Mahalanobis distance.

An essential feature of remote sensing and digital photogrammetric processes is image compression and communication over digital links. This paper investigates the probability of using a convolution-deconvolution method as a pre-post-processing step in standard digital image compression and restoration. As such, the paper relates to image coding and compression systems whereby an original image can be transmitted or stored in a convolved (i.e. blurred) representation which renders it more compressible. The image is then thoroughly restored to its original state by reversing the convolution process.

The compressibility of an image increases with blurring, whereby the relation between the compression ratio (CR) and the blurring scale is almost linear. Hence, by convolving by way of a localised response function (i.e. a linear kernel) and thereby blurring an image before compression, the CR will increase accordingly. In this novel process the response function is applied to a fractal one-dimensional representation of a given image. A blurred image is thus created, which can be shown to contain the details of the original image and thereby restored by reversing the blurring process. The implications of increased CR are examined in terms of the quality of the reconstructed images.

Effective and dynamic recognition of winter wheat has important implications for the development of agriculture in In this paper, we proposed a method for winter wheat identification using particle swarm optimization-support vector machine (PSO-SVM) model and multi-temporal Sentinel-2A image. The eigenvector combination based on spectral information and the eigenvector combination based on texture information were constructed by using different phenological periods of winter wheat. The winter wheat was identified and extracted by PSO-SVM. The extraction accuracy under different feature band combinations was compared and analyzed. The results showed that PSO-SVM had higher accuracy than traditional SVM. Using PSO-SVM, the optimal combination was multi-temporal spectral and mean texture information combination and its classification accuracy was 91.25%. This paper provides a theoretical basis for the future use of Sentinel-2A data to extract other crop information.

In view of the confusing problem of urban river network water and building shadows in hyperspectral images, we analyzed typical shadow and water spectrum in AISA hyperspectral image. On the basis of Normalized Difference Vegetation Index (NDVI), the 588 nm height factor was introduced to constitute an anti-shadow water extraction method (ASWEM). Compared with NDVI extraction results, this method can effectively suppress shadows, especially those cast in buildings, improve water extraction accuracy and reduce water body commission error. The commission error is reduced from 45% to 10.4%, and Kappa coefficient is increased from 0.664 to 0.863. The change of spatial scale has a significant impact on the water extraction results. The lower the image resolution, the more serious the water leakage is, and some small rivers will not be able to extract. However, due to the influence of the mixed pixels, the spectral characteristics of the shadows are weakened to some extent, and the commission error is reduced. As the resolution decreases further, the number and mixing of mixed pixels increases, and the commission error increases.

The linear polarization of light reflected from soil surfaces was measured by an instrument composed of a semi-automatic goniometer and an ASD spectroradiometer under a direct lamp to determine its potential to detect differences in different particle size. In this paper we tested and analyzed the polarization spectra of soils to determine the spectral response and changes in soil particle size, and to establish models of the relationship between spectral data and soil particle size. An orthogonal test was also designed for the various factors that affect soil spectral polarization characteristics and their interactions. All above measurements were carried out in the laboratory where the atmospheric contribution was ignored. The results show that particle size is one of the most important parameter affecting soil spectra, and is critical to soil remote sensing band selection and image interpretation. It also provides information required for soil investigation and analysis of physical and chemical properties.

This paper presents the results of an aerial survey taken over prehistoric and Roman era archaeological sites in southern Moravia and north-west Bohemia. The images were acquired by both aircraft and RPAS, equipped with a variety of sensors (such as multispectral, RGB, infrared, red-edge and thermal cameras), to serve as a source of information for archaeologists about the existence of unknown sites (hitherto not evidenced). The non-invasive character of remotely sensed data which brings important data on the buried archaeological features is valuable and, at the same time, more careful to archaeological heritage in contrast to traditional excavation methods.

Remote sensing is rapid and effective in monitoring crop fields to provide decision support to crop production management in field planning, nutrient management, pest control, irrigation, and harvest. Multi-source, multi-scale, multi-temporal agricultural remote sensing and monitoring provides data with huge volume and high complexity for various analytical applications for effective precision agricultural operations. In the past decade, precision agricultural research have been conducted with the images acquired in the research farms over an area of 400 ha in the center of the Mississippi Delta. The images were acquired from high-resolution satellites, an agricultural airplane, and unmanned aerial vehicles along with ground-based detection and measurement. The image sensors are red-green-blue color, visible-near infrared (VNIR) multispectral, VNIR hyperspectral, and thermal infrared. The image data are not only valuable in research for precision agriculture, weed science, and crop genetics but also able to provide guides for farm consultants and producers in their digital agriculture practices in this area. The purpose of this project is to design and develop a systematic prototype to manage and publish the remote sensing image data acquired from different sources at different spatial and temporal scales on internet and mobile platforms to provide services to the local, regional, national, and even global professionals and farmers. To accommodate all data products, the images have to be resampled to fit into a global image tile structure with a data cube by stacking the image tiles in time sequences covering the same area on the ground. The application of a global image tile structure allows the local data tied into a global remote sensing big data management framework.

Rocky desertification is one of the most serious problems in environmental deterioration, and its accurate mapping is of many implications for maintaining Earth’s sustainability. Compared to traditional field surveying approaches, remote sensing (RS), particularly hyperspectral RS, has proved to be a more efficient solution plan. Yet, hyperspectral RS also suffers from the problem of spectral mixing, since rocky desertification may correspond to various fractions of vegetation, bare soil, and exposed rock. Although linear spectral unmixing (LSU) is an effective method for resolving their ratios, different constraint conditions involving the selections of pure objects and end-member spectra may influence the result. To better overcome this basic problem, the key point of parameter comparison for LSU in field hyperspectral sampling of rocky desertification was investigated. In the typical rocky desertification areas in southwest China, an experiment of field hyperspectral RS sampling various scenarios of ground object mixing was conducted. Then, LSU was operated, by firstly classifying the digital photos of the sample plots to derive the proportion of each object. The specific LSU operations were classified into four types. The selection of end-member spectra were classified into three kinds of cases. With the 12 combination cases of the above-listed scenarios compared, we found that the results under the conditions of ANC and full constraint were better than the ASC and unconstrained conditions, and the performance for the end-member selection scenarios from case A to case C was dropping but could handle more complex situations. These inferences can supply a more solid theoretical basis for better implementing spectral unmixing in hyperspectral RS of rocky desertification.

The FY-3C MWHS-2 (Microwave Humidity Sounder-2) data have been assimilated in the ECMWF operational forecasting system showing small positive impacts on short-range forecasts. To assess the impacts of all-sky (i.e. clear, cloudy and precipitating) assimilation of MWHS-2 183 GHz channels on the analysis and forecasts of typhoon precipitation caused by Typhoon Nida which made landfall in Guangdong Province, China in 2016, three experiments (without MWHS-2 data, with MWHS-2 data in clear-sky conditions and with MWHS-2 data in all-sky conditions) have been carried out. RTTOV-SCATT, a fast Radiative Transfer Model for simulating cloud- and precipitation- affected microwave radiances, and a symmetric observation error model for all-sky radiance assimilation are implemented within the Weather Research and Forecasting model data assimilation system (WRFDA) and its three-dimension variational data assimilation scheme is used for all experiments. Compared with the experiment run without MWHS-2 data, the assimilation of MWHS-2 observations in clear-sky and all-sky conditions makes more accurate precipitation forecasts over the north-central Guangdong. Furthermore, compared with the clear-sky experiment, the precipitation distribution over the east of the Pearl River and the rainfall amount in the eastern Guangdong both are more close to the observations due to the improved temperature and humidity analysis by assimilating more cloud- and precipitation- affected radiances over the northern South China Sea and Typhoon Nida in all-sky experiment. Since the results in this study are very encouraging, more experiments are to be run to verify the positive impacts of all-sky assimilation on the prediction of severe weather processes.

While optical laser range finders use random signals to determine distance, a laser diode’s fast frequency noise can perform the task. Moreover, this signal can be applied to physical-random number generation. This research describes a method, whereby laser diode’s frequency noise characteristics generate a large number of physical-random numbers and determine the distance to a target [1] [2]. We tested the random number generating- and distance- measuring capabilities of two types of lasers; a Fabry-Perot-LD and VCSEL: (Vertical Cavity Surface Emitting Laser). With the Fabry-Perot etalon functioning as frequency discriminator, we investigated the physical-random numbers’ characteristics from both Fabry-Perot-LD’s and the VCSEL’s characteristic’s points of view. We verified the generated binary number’s randomness, using NIST FIPS140-2 test, and noted the Random Number Generation (RNG) speed of a FP-LD was 48 Gbit/s, and that of a VCSEL was 159 Gbit/s. When the generation speed of the physical-random number is high, we can increase the sampling rate of our range finders and improve resolution.

In this research, the ground based hyperspectral reflectance of Chinese cabbage and radish depending on the vegetation growth stages was compared to each other. The classifiers namely decision tree, random forest and support vector machine were tested to check the feasibility of classification depending on the difference in hyperspectral reflectance. The ability of classifier was compared with the overall accuracy and kappa coefficient depending on the vegetation growth stages. The spectral merging was applied to find out the optimal spectral bands to make new multispectral sensor based on the commercial band pass filter with full width at half maximum (FWHM) such as 10nm, 25nm, 40nm, 50nm and 80nm. It was ascertained that the pattern of hyperspectral reflectance varied in Chinese cabbage and radish and also found a certain disparity of pattern in different vegetation growing stage. Although the classifying ability of support vector machine with linear method was higher than the other six methods, it was not suitable for new multispectral sensor. Hence, the decision tree with Rpart method is advantageous as a best classifier to make new multispectral sensor in order to separate the hyperspectral reflectance of Chinese cabbage and radish depending on the vegetation growth stages. The substantiates two alternative aggregate of bands 410nm, 430nm, 700nm and 720nm with 10nm of FWHM or 410nm, 440nm, 690nm and 720nm with 25nm of FWHM were suggested to be the best combinations to make new multispectral sensor without the overlap of FWHM.

In the last decade, technologies of unmanned aerial vehicles (UAVs) and small imaging sensors have achieved a significant improvement in terms of equipment cost, operation cost and image quality. These low-cost platforms provide flexible access to high resolution visible and multispectral images. As a result, many studies have been conducted regarding the applications in precision agriculture, such as water stress detection, nutrient status detection, yield prediction, etc. Different from traditional satellite low-resolution images, high-resolution UAVbased images allow much more freedom in image post-processing. For example, the very first procedure in post-processing is pixel classification, or image segmentation for extracting region of interest(ROI). With the very high resolution, it becomes possible to classify pixels from a UAV-based image, yet it is still a challenge to conduct pixel classification using traditional remote sensing features such as vegetation indices (VIs), especially considering various changes during the growing season such as light intensity, crop size, crop color etc. Thanks to the development of deep learning technologies, it provides a general framework to solve this problem. In this study, we proposed to use deep learning methods to conduct image segmentation. We created our data set of pomegranate trees by flying an off-shelf commercial camera at 30 meters above the ground around noon, during the whole growing season from the beginning of April to the middle of October 2017. We then trained and tested two convolutional network based methods U-Net and Mask R-CNN using this data set. Finally, we compared their performances with our dataset aerial images of pomegranate trees. [Tiebiao- add a sentence to summarize the findings and their implications to precision agriculture]