Over the past several decades, push-broom imaging spectrometers have become a common Earth observation tool. Instruments of this type must be calibrated to convert the raw sensor data into units of spectral radiance. Calibration is in this case a two-step process: First, a sensor model is obtained by performing calibration measurements, which is then used to convert raw signals to spectral radiance data. Further processing steps can be performed to correct for optical image distortions. In this work, we show the complete calibration process for push-broom imaging spectrometers, including uncertainty propagation. Although the focus is specifically on calibrating a HySpex VNIR-1600 airborne-imaging spectrometer, all methods can be adapted for other instruments. We discuss the theory of push-broom imaging spectrometers by introducing a generic sensor model, which includes the main parameters and effects of such instruments. Calibrating detector-related effects, such as dark signal, the noise as a function of the signal, and temperature effects is shown. Correcting temperature effects significantly reduces measurement errors. To determine the signal non-linearity, we built a setup based on the light-addition method and improved this method to allow smaller signal level distances of the sampling points of the non-linearity curve. In addition, we investigate the non-linearity of the integration time. The signal (<=15%) and the integration time (<=0.5%) non-linearities can be corrected with negligible errors. After correcting both non-linearity effects, a smearing effect is revealed, which is investigated in detail. We use a collimator and monochromator setup for calibrating the geometric and spectral parameters, respectively. To accurately model the angular and spectral response functions, we propose using cubic splines, which leads to significant improvements compared to previously used Gaussian functions. We present a new method that allows interpolation of the cubic spline based response functions for pixels not measured. The results show that the spectral and geometric properties are non-uniform and change rapidly within a few pixels. The absolute radiometric calibration is performed with a lamp-plaque setup and an integrating sphere is used for flat-fielding. To mitigate the influence of sphere non-uniformities, we rotate the instrument along the across-track angle to measure the same spot of the sphere with each pixel. We investigate potential systematic errors and use Monte Carlo simulations to determine the uncertainties of the radiometric calibration. In addition, we measure the polarization sensitivity with a wire-grid polarizer. Finally, we propose a novel image transformation method that allows manipulation of geometric and spectral properties of each pixel individually. Image distortions can be corrected by changing a pixel's center angles, center wavelength, and response function shape. This is done by using a transformation matrix that maps each pixel of a target sensor B to the pixels of a source sensor A. This matrix is derived from two cross-correlation matrices: Sensor A and itself, and sensor B and sensor A. We provide the mathematical background and discuss the propagation of uncertainty. A case study shows that the method can significantly improve data quality.

Tesis por compendio / [EN] This work is focused on studying of hyperspectral and structured light based tridimensional image analysis about their application on quality and process control of cereal flour industry and derived products. The structured light based tridimensional image analysis has been used to develop a bread dough dynamic fermentation control system. Descriptors obtained from dough shape evolution were used to describe differences between wheat flour batches during fermentation process. In the same way, that system was used to characterize the effect of new ingredients on fermentation process. Those behaviors were analyzed joint to the intern structure of dough during the process, establishing relationships between it and the tridimensional information. Differences in fermentation process were also studied using hyperspectral image analysis. Flours were analyzed using the obtained diffuse reflectance spectra, which contained information within 400-1000 nm of wavelength range. Differences in several spectral bands were correlated with fundamental components of flours such as gluten. That spectral characterization of flours was used to detect adulterations with different grains. Adulterations until 2, 5% of oat, sorghum and corn were detected both flour and bread crumb. The hyperspectral image analysis was also used to control the heat treatment of wheat and oat flours, where spectral information was related to properties of end products. / [ES] El presente trabajo está centrado en el estudio de los sistemas de análisis de imagen hiperespectral y tridimensional basado en luz estructurada para su aplicación en el control de procesos y calidad de la industria harinera y de sus derivados. El sistema de imagen tridimensional basado en luz estructurada fue la base para el desarrollo de un sistema de monitorización en continuo de la fase de fermentación de masas panarias. A partir de descriptores desarrollados relacionados con la variación de la forma del producto durante la operación, se establecieron diferencias entre lotes de harinas de trigo y describió el comportamiento de masas reformuladas con nuevos ingredientes. Dicho comportamiento fue analizado en conjunto con la estructura interna de la masa durante la operación, estableciendo relaciones concretas entre esta y la información obtenida a partir de las imágenes. Las diferencias de comportamiento durante la operación de fermentación también fueron estudiadas mediante el sistema de imagen hiperespectral. En este caso, las harinas fueron analizadas directamente mediante imágenes espectrales, obteniendo espectros de reflectancia difusa en el rango de longitudes de onda 400-1000, donde se observaron diferencias en determinadas bandas del espectro. Dichas diferencias fueron correlacionadas con determinados componentes fundamentales como el gluten. La caracterización espectral de la harina de trigo se utilizó posteriormente para la detección de cereales diferentes mezclados con esta. Adulteraciones de hasta un 2,5% de avena, sorgo y maíz fueron detectadas tanto en harina como en panes de trigo. El análisis de imagen hiperespectral también ha sido aplicado al control del tratamiento térmico de harinas de trigo y avena, donde se ha conseguido relacionar la información espectral con las características del producto final. / [CA] El present treball està centrat en l'estudi dels sistemes d'anàlisis d'imatge hiperespectral i tridimensional basat en llum estructurada per a la seua aplicació en el control de processos i qualitat de la indústria farinera i dels seus derivats. El sistema d'imatge tridimensional basat en llum estructurada va ser la base per al desenvolupament d'un sistema de monitoratge en continu de la fase de fermentació de masses panaries. A partir dels descriptors desenvolupats relacionats amb la variació de la forma del producte durant l'operació, es van establir diferències entre lots de farines de blat i es va descriure el comportament de masses reformulades amb nous ingredients. Aquest comportament va ser analitzat en conjunt amb l'estructura interna de la massa durant l'operació, establint relacions concretes entre aquesta i la informació obtinguda a partir de les imatges. Les diferències de comportament durant l'operació de fermentació també van ser estudiades mitjançant el sistema d'imatge hiperespectral. En aquest cas, les farines van ser analitzades directament mitjançant imatges espectrals, obtenint espectres de reflectància difusa en el rang de longituds d'ona 400-1000, on es van observar diferències en determinades bandes de l'espectre. Aquestes diferències van ser correlacionades amb determinats components fonamentals com el gluten. La caracterització espectral de la farina de blat es va utilitzar posteriorment per a la detecció de cereals diferents barrejats amb aquesta. Adulteracions de fins a un 2,5% de civada, sorgo i dacsa van ser detectades tant en farina com en pans de blat. L'anàlisi d'imatge hiperespectral també ha sigut aplicat al control del tractament tèrmic de farines de blat i civada, on s'ha aconseguit relacionar la informació espectral amb les característiques del producte final. / Verdú Amat, S. (2016). Aplicación del análisis de imagen hiperespectral y tridimensional al control de procesos y productos en la industria harinera y sus derivados [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/65354 / TESIS / Premios Extraordinarios de tesis doctorales / Compendio

Indiana University-Purdue University Indianapolis (IUPUI) / Water is an essential resource for life on Earth, and monitoring its quality is an important task for mankind. However, the amount of water quality data collected by the traditional method is insufficient for the conservation and sustainable management of this important resource. This challenge will be exacerbated by increasing harmful algal blooms at the global scale. To fill this gap, Earth Observations (EO) have been proposed to help stakeholders make their decisions, but the use of EO for monitoring inland water quality is still in development. In this context, the main objective of this study was to improve the estimation of cyanobacteria via remote sensing data. To achieve this goal, the water type classification was first used to identify the dominant optically active constituents within aquatic environments. This information is crucial for understanding the optical properties of inland waters and selecting the best remote sensing algorithm for specific optical water types. The next research question was to develop a universal structure for retrieval of the inherent optical properties of several important aquatic systems around the world, which can be used as a corner stone for developing a globally applicable remote sensing algorithm. The third research topic of this dissertation is about removing the interference of chlorophyll-a with the absorption strength at 620 nm where phycocyanin exhibits its diagnostic absorption so that the estimation of phycocyanin concentration can be improved. Despite the novelty of the proposed remote sensing algorithms which are able to accommodate distinct water optical properties, there are abundant opportunities for improving the parameterization of the proposed models to retrieve inland water quality and optical properties when a global database of optical and water quality measurements is available. Considering the current advancement in spaceborne technology and the existence of a coordinate effort for global calibration and validation of remote sensing algorithms for monitoring inland waters, there is a high potential for operational assessment of harmful cyanobacterial blooms using the remote sensing algorithms proposed in this dissertation.

In this dissertation, novel techniques for hyperspectral classification and signal reconstruction from random projections are presented. A classification paradigm designed to exploit the rich statistical structure of hyperspectral data is proposed. The proposed framework employs the local Fisher’s discriminant analysis to reduce the dimensionality of the data while preserving its multimodal structure, followed by a subsequent Gaussianmixture- model or support-vector-machine classifier. An extension of this framework in a kernel induced space is also studied. This classification approach employs a maximum likelihood classifier and dimensionality reduction based on a kernel local Fisher’s discriminant analysis. The technique imposes an additional constraint on the kernel mapping—it ensures that neighboring points in the input space stay close-by in the projected subspace. In a typical remote sensing flow, the sender needs to invoke an appropriate compression strategy for downlinking signals (e.g., imagery to a base station). Signal acquisition using random projections significantly decreases the sender-side computational cost, while preserving useful information. In this dissertation, a novel class-dependent hyperspectral image reconstruction strategy is also proposed. The proposed method employs statistics pertinent to each class as opposed to the average statistics estimated over the entire dataset, resulting in a more accurate reconstruction from random projections. An integrated spectral-spatial model for signal reconstruction from random projections is also developed. In this approach, spatially homogeneous segments are combined with spectral pixel-wise classification results in the projected subspace. An appropriate reconstruction strategy, such as compressive projection principal component analysis (CPPCA), is employed individually in each category based on this integrated map. The proposed method provides better reconstruction performance as compared to traditional methods and the class-dependent CPPCA approach.

The High Efficiency HyperSpectral Imager (HEHSI) is a Fourier Transform hyperspectral imager based on a Sagnac interferometer. This thesis research concentrates on the design upgrade and calibration of HEHSI from a proof of concept instrument to a prototype field instrument. Stability is enhanced by removing degrees of freedom and alignment is enhanced by providing for in-situ adjustments. The use of off the shelf components allows for reduced development time and cost constraints. HEHSI is capable of multiple configurations to accommodate sensors and optics with specialized capabilities for multiple wavelength ranges and viewing conditions. With a spectral response of 400 to 1000 nanometers in the visible and very near IR as well as 900 to 1700nm in the Near IR. Creation and use of a real time feedback alignment utility allow quantifiable signal comparison and image alignment. Advances allow for HEHSI to remain aligned during data collection sessions and confirmation of alignment through quantitative measures.

Hyperspectral imagery is a highly dimensional type of data resulting in high computational costs during analysis. Band selection aims to reduce the original hyperspectral image to a smaller subset that reduces these costs while preserving the maximum amount of spectral information within the data. This thesis explores various types of band selection techniques used in hyperspectral image processing. Modifying Neural network-based techniques and observing the effects on the band selection process due to the change in network architecture or objective are of particular focus in this thesis. Herein, a generalized neural network-based band selection technique is developed and compared to state-of-the-art algorithms that are applied to a unique dataset and the Pavia City Center dataset where the subsequent selected bands are fed into a classifier to gather comparison results.

It is a popular practice in the remote-sensing community to apply principal component analysis (PCA) on a higher-dimensional feature space to achieve dimensionality-reduction. Several factors that have led to the popularity of PCA include its simplicity, ease of use, availability as part of popular remote-sensing packages, and optimal nature in terms of mean square error. These advantages have prompted the remote-sensing research community to overlook many limitations of PCA when used as a dimensionality-reduction tool for classification and target-detection applications. This thesis addresses the limitations of PCA when used as a dimensionality-reduction technique for extracting discriminating features from hyperspectral data. Theoretical and experimental analyses are presented to demonstrate that PCA is not necessarily an appropriate feature-extraction method for high-dimensional data when the objective is classification or target-recognition. The influence of certain data-distribution characteristics, such as within-class covariance, between-class covariance, and correlation on PCA transformation, is analyzed in this thesis. The classification accuracies obtained using PCA features are compared to accuracies obtained using other feature-extraction methods like variants of Karhunen-Loève transform and greedy search algorithms on spectral and wavelet domains. Experimental analyses are conducted for both two-class and multi-class cases. The classification accuracies obtained from higher-order PCA components are compared to the classification accuracies of features extracted from different regions of the spectrum. The comparative study done on the classification accuracies that are obtained using above feature-extraction methods, ascertain that PCA may not be an appropriate tool for dimensionality-reduction of certain hyperspectral data-distributions, when the objective is classification or target-recognition.

Influence of nitrogen (N) deficiency on castor cv. ‘Hale’ growth, physiology, and leaf reflectance properties were investigated. Treatments imposed were complete Hoagland’s nutrient solution (control, 100N), reduced N to 20% of the control (20N) and withheld N from the solution (0N) from 34 to 66 days after sowing (DAS) in 12-L pots grown out doors. N deficiency significantly reduced leaf area, chlorophyll and photosynthesis resulting lower total biomass. Leaf and stem growth rates were more sensitive to leaf N concentration than photosynthesis and leaf addition rates. N deficiency stress increased leaf reflectance at R555 and R715 nm and caused a red-edge shift to shorter wavelengths. Reflectance ratios of R455/R605 and R505/R605 nm was highly correlated with leaf N on weight (r2 = 0.93) and area-based (r2 = 0.90) estimations, respectively. Similarly, reflectance ratio R635/R505 was highly correlated with chlorophylls (r2 = 0.94). The N-specific wavebands and functional relationships between leaf N and growth and developmental processes would be useful for rapid and non-destructive estimation of leaf N and growth rates of castor.

No description available.

No description available.