This PDF file contains the front matter associated with SPIE Proceedings Volume 10213, including the Title Page, Copyright information, Table of Contents, and Conference Committee listing.

Hyperspectral imaging (HSI) has been used for over two decades in laboratory research, academic, environmental and defense applications. In more recent time, HSI has started to be adopted for commercial applications in machine vision, conservation, resource exploration, and precision agriculture, to name just a few of the economically viable uses for the technology. Corning Incorporated (Corning) has been developing and manufacturing HSI sensors, sensor systems, and sensor optical engines, as well as HSI sensor components such as gratings and slits for over a decade and a half. This depth of experience and technological breadth has allowed Corning to design and develop unique HSI spectrometers with an unprecedented combination of high performance, low cost and low Size, Weight, and Power (SWaP). These sensors and sensor systems are offered with wavelength coverage ranges from the visible to the Long Wave Infrared (LWIR). The extremely low SWaP of Corning’s HSI sensors and sensor systems enables their deployment using limited payload platforms such as small unmanned aerial vehicles (UAVs).

This paper discusses use of the Corning patented monolithic design Offner spectrometer, the microHSI™, to build a highly compact 400-1000 nm HSI sensor in combination with a small Inertial Navigation System (INS) and micro-computer to make a complete turn-key airborne remote sensing payload. This Selectable Hyperspectral Airborne Remote sensing Kit (SHARK) has industry leading SWaP (1.5 lbs) at a disruptively low price due, in large part, to Corning’s ability to manufacture the monolithic spectrometer out of polymers (i.e. plastic) and therefore reduce manufacturing costs considerably. The other factor in lowering costs is Corning’s well established in house manufacturing capability in optical components and sensors that further enable cost-effective fabrication. The competitive SWaP and low cost of the microHSI™ sensor is approaching, and in some cases less than the price point of Multi Spectral Imaging (MSI) sensors. Specific designs of the Corning microHSI™ SHARK visNIR turn-key system are presented along with salient performance characteristics. Initial focus market areas include precision agriculture and historic and recent microHSI™ SHARK prototype test results are presented.

Hyperspectral imaging can provide hundreds of images at different wave bands covering the visible and near infrared regions, which is superior to traditional spectral and RGB techniques. Minnesota produced a lot of maize every year, while the temperature in Minnesota can change abruptly during spring. This study was carried out to use hyperspectral imaging technique to identify maize seedlings with cold stress prior to having visible phenotypes. A total of 60 samples were scanned by the hyperspectral camera at the wave range of 395-885 nm. The spectral reflectance information was extracted from the corrected hyperspectral images. By spectral reflectance information, support vector machine (SVM) classification models were established to identify the cold stressed samples. Then, the wavelengths which could play significant roles for the detection were selected using two-wavelength combination method. The classifiers were built again using the selected wavelengths. From the results, it can be found the selected wavelengths can even perform better than full wave range. The overall results indicated that hyperspectral imaging has the potential to classify cold stress symptoms in maize seedlings and thus help in selecting the corn genome lines with cold stress resistance.

A flat-field aberration corrected concave blaze grating for 400-1100nm is designed and fabricated. The concave grating, which has a free-form profile with blaze grating pitch and variable line spacing, is fabricated using five-axis CNC machine with nanometer machining precision. An optical system is setup to measure the focused spot size, spectral resolution and diffraction efficiency to evaluate the performance of aberration-corrected concave grating. The blaze grating reaches a diffraction efficiency of 70%. The focused vertical spot size is 50µm, which indicates a 50µm spatial resolution at the image plane. The focused horizontal spot size is 300µm, which converts to a spectral resolution of 6nm. The design methodology can be applied to an Offner type hyperspectral imager with a free-form convex grating and variable line spacing to achieve high efficiency and high spatial and high spectral resolving power.

Polyunsaturated fatty acids (PUFAs) are important nutrients present in Salmon. However, current methods for quantifying the fatty acids (FAs) contents in foods are generally based on gas chromatography (GC) technique, which is time-consuming, laborious and destructive to the tested samples. Therefore, the capability of near-infrared (NIR) hyperspectral imaging to predict the PUFAs contents of C20:2 n-6, C20:3 n-6, C20:5 n-3, C22:5 n-3 and C22:6 n-3 in Salmon fillets in a rapid and non-destructive way was investigated in this work. Mean reflectance spectra were first extracted from the region of interests (ROIs), and then the spectral pre-processing methods of 2^nd derivative and Savitzky-Golay (SG) smoothing were performed on the original spectra. Based on the original and the pre-processed spectra, PLSR technique was employed to develop the quantitative models for predicting each PUFA content in Salmon fillets. The results showed that for all the studied PUFAs, the quantitative models developed using the pre-processed reflectance spectra by “2^nd derivative + SG smoothing” could improve their modeling results. Good prediction results were achieved with RP and RMSEP of 0.91 and 0.75 mg/g dry weight, 0.86 and 1.44 mg/g dry weight, 0.82 and 3.01 mg/g dry weight for C20:3 n-6, C22:5 n-3 and C20:5 n-3, respectively after pre-processing by “2^nd derivative + SG smoothing”. The work demonstrated that NIR hyperspectral imaging could be a useful tool for rapid and non-destructive determination of the PUFA contents in fish fillets.

This presentation summarizes the advances made over 40+ years in understanding, modeling, and mapping terrestrial vegetation as reported in the new book on “Hyperspectral Remote Sensing of Vegetation” (Publisher:Taylor and Francis inc.). The advent of spaceborne hyperspectral sensors or imaging spectroscopy (e.g., NASA’s Hyperion, ESA’s PROBA, and upcoming Italy’s ASI’s Prisma, Germany’s DLR’s EnMAP, Japanese HIUSI, NASA’s HyspIRI) as well as the advances made in processing when handling large volumes of hyperspectral data have generated tremendous interest in advancing the hyperspectral applications’ knowledge base to large areas. Advances made in using hyperspectral data, relative to broadband data, include: (a) significantly improved characterization and modeling of a wide array of biophysical and biochemical properties of vegetation, (b) ability to discriminate plant species and vegetation types with high degree of accuracy, (c) reducing uncertainties in determining net primary productivity or carbon assessments from terrestrial vegetation, (d) improved crop productivity and water productivity models, (e) ability to assess stress resulting from causes such as management practices, pests and disease, water deficit or water excess, and (f) establishing more sensitive wavebands and indices to study vegetation characteristics. The presentation will discuss topics such as: (1) hyperspectral sensors and their characteristics, (2) methods of overcoming the Hughes phenomenon, (3) characterizing biophysical and biochemical properties, (4) advances made in using hyperspectral data in modeling evapotranspiration or actual water use by plants, (5) study of phenology, light use efficiency, and gross primary productivity, (5) improved accuracies in species identification and land cover classifications, and (6) applications in precision farming.

A hyperspectral image provides multidimensional figure rich in data consisting of hundreds of spectral dimensions. Analyzing the spectral and spatial information of such image with linear and non-linear algorithms will result in high computational time. In order to overcome this problem, this research presents a system using a MapReduce-Graphics Processing Unit (GPU) model that can help analyzing a hyperspectral image through the usage of parallel hardware and a parallel programming model, which will be simpler to handle compared to other low-level parallel programming models. Additionally, Hadoop was used as an open-source version of the MapReduce parallel programming model. This research compared classification accuracy results and timing results between the Hadoop and GPU system and tested it against the following test cases: the CPU and GPU test case, a CPU test case and a test case where no dimensional reduction was applied.

This paper reports the development of 2000×256 format SWIR HgCdTe/Si FPA with multiple-choice gain (i.e. multiple-choice charge handling capacity) for hyperspectral detection. The spectral resolution is about 8nm. To meet the demands of variable low flux detection within each spectral band in the short wave infrared range, low dark current, low noise, variable conversion gains and high SNR (Signal to Noise Ratio) of FPA are needed. In this paper, we fabricate 512×512 pixel 30μm pitch SWIR HgCdTe diode array on Si by using a novel stress-release construction of HgCdTe chip on Si. Moreover, we design low noise, variable conversion gain and large dynamic range read-out integrated circuit (ROIC) and hybridized the ROIC on the HgCdTe diode array on Si substrate. There are 8-choice gains which can be selected locally according to the incident flux to meet high SNR detection demand. By high-accuracy splicing 4 512×512 HgCdTe/Si FPA we get mosaic 2000×512 FPA, and characterizations have been carried out and reveal that the array dark current densities on an order of 10^-10A/cm², quantum efficiency exceeding 70%, and the operability of 99.5% at operating temperature of around 110K. The SNR of this FPA achieved 120 when illuminated under 5×10⁴photons/pixel.

Combining reflectometry and hyperspectral imaging allows mapping of thin film thickness. Therefore, layer thickness is calculated by comparing a dataset of simulated spectra with the measured data. Utilizing the maximum frame rate of the hyperspectral imager, the pixel wise spectra comparing procedure cannot be performed using a standard computer due to the processing load. In this work, a method using neural networks for calculating layer thickness is presented. By the use of the nonlinear equation as result of a trained neural network, thickness data can be determined with a measurement rate matching the maximum frame rate of the hyperspectral imager.

The current state of the art of specifying spectral imagers falls short of what is needed. Commercial datasheets do not adequately reflect the performance of hyperspectral imagers offered as standard products. In particular, imperfections such as coregistration error, noise performance and stray light are rarely well specified. A standardized way to specify spectral imagers would benefit both developers and users of such instruments. The paper reviews the many different characteristics that are needed to describe various aspects of imager performance, and discusses possible ways to form figures of merit relevant to application performance. In particular, the product of quantum efficiency, optics transmission and nominal throughput (étendue) is shown to be a good figure of merit for radiometric performance. A list of about 30 characteristics is suggested as a standard for a complete specification of spectral imagers. For some characteristics, notably coregistration, it is necessary to establish a standardized measurement methodology.

Precision agriculture can be considered as one of the solutions to optimize agricultural practice such as nitrogen fertilization. Nitrogen deficiency is a major limitation to crop production worldwide whereas excess leads to environmental pollution. In this context, some devices were developed as reflectance spot sensors for on-the-go applications to detect leaves nitrogen concentration deduced from chlorophyll concentration. However, such measurements suffer from interferences with the crop growth stage and the water content of plants. The aim of this contribution is to evaluate the nitrogen status in winter wheat by using multispectral imaging. The proposed system is composed of a CMOS camera and a set of filters ranged from 450 nm to 950 nm and mounted on a wheel which moves due to a stepper motor. To avoid the natural irradiance variability, a white reference is used to adjust the integration time. The segmentation of Photosynthetically Active Leaves is performed by using Bayes theorem to extract their mean reflectance. In order to introduce information related to the canopy architecture, i.e. the crop growth stage, textural attributes are also extracted from raw images at different wavelength ranges. N_c was estimated by partial least squares regression (R^² = 0.94). The best attribute was homogeneity extracted from the gray level co-occurrence matrix (R^² = 0.91). In order to select in limited number of filters, best subset selection was performed. N_c could be estimated by four filters (450 ± 40 nm, 500 ± 20 nm, 650 ± 40 nm, 800 ± 50 nm) (R^² = 0.91).

In this study we attempt to identify the most suitable spectral bands to discriminate among wheat and oat crops using field hyperspectral remote sensing. Discrimination of these crops using ordinary aerial or multispectral satellite imagery can be challenging. Even though multispectral images could have a high spatial resolution, their few wide spectral bands hinder crop discrimination. Therefore, both high spatial resolution and spectral resolution are necessary to accurately discriminate between visually similar crops. One field each of oats and spring wheat, each at least 10 acres in size, was selected in southeastern Wisconsin. Biweekly spectral readings were taken using a spectroradiometer during the growing season from May to July. In each field, seven 10 m x 10 m quadrants were randomly placed and in each quadrants five points were selected from which 20 radiometric readings were taken. Radiometric measurements taken at each sampling point were averaged to derive a single reflectance curve per sampling date, covering the spectral range of 300 nm to 2,500 nm. Each spectral curve was divided into hyperspectral bands each 3 nm wide. The Mann-Whitney U-test was used to estimate how separable the two crops were. Results show that selected regions of the visible light and infrared radiation spectrum have the potential to discriminate between these crops. Crop discrimination is one of the first steps to support crop monitoring and agricultural surveys efforts.

A four cable-driven parallel manipulator (CDPM), consisting of sophisticated spectrometers and imagers, is under development for use in acquiring phenotypic and environmental data over an acre-sized crop field. To obtain accurate and high quality data from the instruments, the end effector must be stable during sensing. One of the factors that reduces stability is the center of mass offset of the end effector, which can cause a pendulum effect or undesired tilt angle. The purpose of this work is to develop a system and method for balancing the center of mass of a 12^th-scale CDPM to minimize vibration that can cause error in the acquired data. A simple method for balancing the end effector is needed to enable end users of the CDPM to arbitrarily add and remove sensors and imagers from the end effector as their experiments may require. A Center of Mass Balancing System (CMBS) is developed in this study which consists of an adjustable system of weights and a gimbal for tilt mitigation. An electronic circuit board including an orientation sensor, wireless data communication, and load cells was designed to validate the CMBS. To measure improvements gained by the CMBS, several static and dynamic experiments are carried out. In the experiments, the dynamic vibrations due to the translational motion and static orientation were measured with and without CMBS use. The results show that the CMBS system improves the stability of the end-effector by decreasing vibration and static tilt angle.

For ecologically valuable areas monitoring, precise agriculture and forestry, thematic maps or small GIS are needed. Remotely Piloted Aircraft Systems (RPAS) data can be obtained on demand in a short time with cm resolution. Data collection is environmentally friendly and low-cost from an economical point of view. This contribution is focused on using eBee drone for mapping or monitoring national natural reserve which is not opened to public and partly pure inaccessible because its moorland nature. Based on a new equipment (thermal imager, multispectral imager, NIR, NIR red-edge and VIS camera) we started new projects in precise agriculture and forestry.