L’imagerie hyperspectrale est un type d’imagerie émergent qui connaît un essor important depuis le début des années 2000. Grâce à une structure spectrale très fine qui produit un volume de donnée très important, elle apporte, par rapport à l’imagerie visible classique, un supplément d’information pouvant être mis à profit dans de nombreux domaines d’exploitation. Nous nous intéressons spécifiquement à la détection et l’analyse de changements entre deux images de la même scène, pour des applications orientées vers la défense.Au sein de ce manuscrit, nous commençons par présenter l’imagerie hyperspectrale et les contraintes associées à son utilisation pour des problématiques de défense. Nous présentons ensuite une méthode de détection et de classification de changements basée sur la recherche de directions spécifiques dans l’espace généré par le couple d’images, puis sur la fusion des directions proches. Nous cherchons ensuite à exploiter l’information obtenue sur les changements en nous intéressant aux possibilités de dé-mélange de séries temporelles d’images d’une même scène. Enfin, nous présentons un certain nombre d’extensions qui pourront être réalisées afin de généraliser ou améliorer les travaux présentés et nous concluons. / Hyperspectral imagery is an emerging imagery technology which has known a growing interest since the 2000’s. This technology allows an impressive growth of the data registered from a specific scene compared to classical RGB imagery. Indeed, although the spatial resolution is significantly lower, the spectral resolution is very small and the covered spectral area is very wide. We focus on change detection between two images of a given scene for defense oriented purposes.In the following, we start by introducing hyperspectral imagery and the specificity of its exploitation for defence purposes. We then present a change detection and analysis method based on the search for specifical directions in the space generated by the image couple, followed by a merging of the nearby directions. We then exploit this information focusing on theunmixing capabilities of multitemporal hyperspectral data. Finally, we will present a range of further works that could be done in relation with our work and conclude about it.

Cette thèse est consacrée au démixage en imagerie hyperspectrale en particulier dans le cas où des objets de petite taille sont présents dans la scène. Les images hyperspectrales contiennent une grande quantité d’information à la fois spectrale et spatiale, et chaque pixel peut être vu comme le spectre de réflexion de la zone imagée. Du fait de la faible résolution spatiale des capteurs le spectre de réflexion observé au niveau de chaque pixel est un mélange des spectres de réflexion de l’ensemble des composants imagés dans le pixel. Une problématique de ces images hyperspectrales est le démixage, qui consiste à décomposer l’image en une liste de spectres sources, appelés endmembers, correspondants aux spectres de réflexions des composants de la scène d’une part, et d’autre part la proportion de chacun de ces spectres source dans chaque pixel de l’image. De nombreuses méthodes de démixage existent mais leur efficacité reste amoindrie en présence de spectres sources dits rares (c’est-à-dire des spectres présents dans très peu de pixels, et souvent à un niveau subpixelique). Ces spectres rares correspondent à des composants présents en faibles quantités dans la scène et peuvent être vus comme des anomalies dont la détection est souvent cruciale pour certaines applications.Nous présentons dans un premier temps deux méthodes de détection des pixels rares dans une image, la première basée sur un seuillage de l’erreur de reconstruction après estimation des endmembers abondants, la seconde basée sur les coefficients de détails obtenus par la décomposition en ondelettes. Nous proposons ensuite une méthode de démixage adaptée au cas où une partie des endmembers sont connus a priori et montrons que cette méthode utilisée avec les méthodes de détection proposées permet le démixage des endmembers des pixels rares. Enfin nous étudions une méthode de rééchantillonnage basée sur la méthode du bootstrap pour amplifier le rôle de ces pixels rares et proposer des méthodes de démixage en présence d’objets de petite taille. / This thesis is devoted to the unmixing issue in hyperspectral images, especiallyin presence of small sized objects. Hyperspectral images contains an importantamount of both spectral and spatial information. Each pixel of the image canbe assimilated to the reflection spectra of the imaged scene. Due to sensors’ lowspatial resolution, the observed spectra are a mixture of the reflection spectraof the different materials present in the pixel. The unmixing issue consists inestimating those materials’ spectra, called endmembers, and their correspondingabundances in each pixel. Numerous unmixing methods have been proposed butthey fail when an endmembers is rare (that is to say an endmember present inonly a few of the pixels). We call rare pixels, pixels containing those endmembers.The presence of those rare endmembers can be seen as anomalies that we want todetect and unmix. In a first time, we present two detection methods to retrievethis anomalies. The first one use a thresholding criterion on the reconstructionerror from estimated dominant endmembers. The second one, is based on wavelettransform. Then we propose an unmixing method adapted when some endmembersare known a priori. This method is then used with the presented detectionmethod to propose an algorithm to unmix the rare pixels’ endmembers. Finally,we study the application of bootstrap resampling method to artificially upsamplerare pixels and propose unmixing methods in presence of small sized targets.

abstract: Hyperspectral unmixing is an important remote sensing task with applications including material identification and analysis. Characteristic spectral features make many pure materials identifiable from their visible-to-infrared spectra, but quantifying their presence within a mixture is a challenging task due to nonlinearities and factors of variation. In this thesis, physics-based approaches are incorporated into an end-to-end spectral unmixing algorithm via differentiable programming. First, sparse regularization and constraints are implemented by adding differentiable penalty terms to a cost function to avoid unrealistic predictions. Secondly, a physics-based dispersion model is introduced to simulate realistic spectral variation, and an efficient method to fit the parameters is presented. Then, this dispersion model is utilized as a generative model within an analysis-by-synthesis spectral unmixing algorithm. Further, a technique for inverse rendering using a convolutional neural network to predict parameters of the generative model is introduced to enhance performance and speed when training data are available. Results achieve state-of-the-art on both infrared and visible-to-near-infrared (VNIR) datasets as compared to baselines, and show promise for the synergy between physics-based models and deep learning in hyperspectral unmixing in the future. / Dissertation/Thesis / Masters Thesis Electrical Engineering 2020

Hyperspectral imagery is often associated with high storage and transmission costs. Dimensionality reduction aims to reduce the time and space complexity of hyperspectral imagery by projecting data into a low-dimensional space such that all the important information in the data is preserved. Dimensionality-reduction methods based on transforms are widely used and give a data-dependent representation that is unfortunately costly to compute. Recently, there has been a growing interest in data-independent representations for dimensionality reduction; of particular prominence are random projections which are attractive due to their computational efficiency and simplicity of implementation. This dissertation concentrates on exploring the realm of computationally fast and efficient random projections by considering projections based on a random Hadamard matrix. These Hadamard-based projections are offered as an alternative to more widely used random projections based on dense Gaussian matrices. Such Hadamard matrices are then coupled with a fast singular value decomposition in order to implement a two-stage dimensionality reduction that marries the computational benefits of the data-independent random projection to the structure-capturing capability of the data-dependent singular value transform. Finally, random projections are applied in conjunction with nonnegative least squares to provide a computationally lightweight methodology for the well-known spectral-unmixing problem. Overall, it is seen that random projections offer a computationally efficient framework for dimensionality reduction that permits hyperspectral-analysis tasks such as unmixing and classification to be conducted in a lower-dimensional space without sacrificing analysis performance while reducing computational costs significantly.

This report presents an investigation and development of the methods for orbital object identification. Two goals were accomplished in this master’s thesis; the development of a method of inverting material proportions from an object’s combined spectrum, and the investigation of methods and initialization of measurement of space environment effects on spectral features of common spacecraft materials. A constrained least squares approach was chosen for inverting spectral proportions from the combined spectra. The final results fall within 1 - 15% of the original spectrum, depending on the quality and noise levels of the original spectrum. Additionally, the effects of outgassing and atomic oxygen erosion were measured using the vacuum chamber facilities at California Polytechnic State University and are to be used as a basis for future identification of orbital debris. To have a fully functional model for accurately identifying space objects, both parts are needed: a set of space environment effect measurements as a basis for the identification model (for use on objects exposed to the space environment), and the identification model to mathematically determine the best fit set of materials.

No description available.

This work aims to develop a deep learning model for land cover quantification through hyperspectral unmixing using an unsupervised autoencoder. Land cover identification and classification is instrumental in urban planning, environmental monitoring and land management. With the technological advancements in remote sensing, hyperspectral imagery which captures high resolution images of the earth's surface across hundreds of wavelength bands, is becoming increasingly popular. The high spectral information in these images can be analyzed to identify the various target materials present in the image scene based on their unique reflectance patterns. An autoencoder is a deep learning model that can perform spectral unmixing by decomposing the complex image spectra into its constituent materials and estimating their abundance compositions. The advantage of using this technique for land cover quantification is that it is completely unsupervised and eliminates the need for labelled data which generally requires years of field survey and formulation of detailed maps. We evaluate the performance of the autoencoder on various synthetic and real hyperspectral images consisting of different land covers using similarity metrics and abundance maps. The scalability of the technique with respect to landscapes is assessed by evaluating its performance on hyperspectral images spanning across 100m x 100m, 200m x 200m, 1000m x 1000m, 4000m x 4000m and 5000m x 5000m regions. Finally, we analyze the performance of this technique by comparing it to several supervised learning methods like Support Vector Machine (SVM), Random Forest (RF) and multilayer perceptron using F1-score, Precision and Recall metrics and other unsupervised techniques like K-Means, N-Findr, and VCA using cosine similarity, mean square error and estimated abundances. The land cover classification obtained using this technique is compared to the existing United States National Land Cover Database (NLCD) classification standard. / Master of Science / This work aims to develop an automated deep learning model for identifying and estimating the composition of the different land covers in a region using hyperspectral remote sensing imagery. With the technological advancements in remote sensing, hyperspectral imagery which captures high resolution images of the earth's surface across hundreds of wavelength bands, is becoming increasingly popular. As every surface has a unique reflectance pattern, the high spectral information contained in these images can be analyzed to identify the various target materials present in the image scene. An autoencoder is a deep learning model that can perform spectral unmixing by decomposing the complex image spectra into its constituent materials and estimate their percent compositions. The advantage of this method in land cover quantification is that it is an unsupervised technique which does not require labelled data which generally requires years of field survey and formulation of detailed maps. The performance of this technique is evaluated on various synthetic and real hyperspectral datasets consisting of different land covers. We assess the scalability of the model by evaluating its performance on images of different sizes spanning over a few hundred square meters to thousands of square meters. Finally, we compare the performance of the autoencoder based approach with other supervised and unsupervised deep learning techniques and with the current land cover classification standard.

L'imagerie hyperspectrale consiste à acquérir une scène spatiale à plusieurs longueurs d'onde, e.g. en microscopie. Cependant, lorsque l'image est observée à une résolution suffisamment fine, elle est dégradée par un flou (convolution) et une procédure de déconvolution doit être utilisée pour restaurer l'image originale. Ce problème inverse, par opposition au problème direct modélisant la dégradation de l'image observée, est étudié dans la première partie . Un autre problème inverse important en imagerie, la séparation de sources, consiste à extraire les spectres des composants purs de l'image (sources) et à estimer les contributions de chaque source à l'image. La deuxième partie propose des contributions algorithmiques en restauration d'images hyperspectrales. Le problème est formulé comme la minimisation d'un critère pénalisé et résolu à l'aide d'une structure de calcul rapide. La méthode est adaptée à la prise en compte de différents a priori sur l'image, tels que sa positivité ou la préservation des contours. Les performances des techniques proposées sont évaluées sur des images de biocapteurs bactériens en microscopie confocale de fluorescence. La troisième partie est axée sur le problème de séparation de sources, abordé dans un cadre géométrique. Nous proposons une nouvelle condition suffisante d'identifiabilité des sources à partir des coefficients de mélange. Une étude innovante couplant le modèle d'observation avec le mélange de sources permet de montrer l'intérêt de la déconvolution comme étape préliminaire de la séparation. Ce couplage est validé sur des données acquises en spectroscopie Raman / Hyperspectral imaging refers to the acquisition of spatial images at many spectral bands, e.g. in microscopy. Processing such data is often challenging due to the blur caused by the observation system, mathematically expressed as a convolution. The operation of deconvolution is thus necessary to restore the original image. Image restoration falls into the class of inverse problems, as opposed to the direct problem which consists in modeling the image degradation process, treated in part 1 of the thesis. Another inverse problem with many applications in hyperspectral imaging consists in extracting the pure materials making up the image, called endmembers, and their fractional contribution to the data or abundances. This problem is termed spectral unmixing and its resolution accounts for the nonnegativity of the endmembers and abundances. Part 2 presents algorithms designed to efficiently solve the hyperspectral image restoration problem, formulated as the minimization of a composite criterion. The methods are based on a common framework allowing to account for several a priori assumptions on the solution, including a nonnegativity constraint and the preservation of edges in the image. The performance of the proposed algorithms are demonstrated on fluorescence confocal images of bacterial biosensors. Part 3 deals with the spectral unmixing problem from a geometrical viewpoint. A sufficient condition on abundance coefficients for the identifiability of endmembers is proposed. We derive and study a joint observation model and mixing model and demonstrate the interest of performing deconvolution as a prior step to spectral unmixing on confocal Raman microscopy data

Dans cette thèse, nous avons présenté les aspects de la technologie d'imagerie hyperspectrale en concentrant sur le problème de démélange non-linéaire. Pour cette tâche, nous avons proposé trois solutions. La première consiste à intégrer les avantages de l'apprentissage de variétés dans les méthodes de démélange classique pour concevoir leurs versions non-linéaires. Les résultats avec les données générées sur une variété bien connue - le "Swissroll"- donne des résultats prometteurs. Les méthodes fonctionnent beaucoup mieux avec l'augmentation de la non-linéarité. Cependant, l'absence de contrainte de non-négativité dans ces méthodes reste une question ouverte pour des améliorations à trouver. La deuxième proposition vise à utiliser la méthode de pré-image pour estimer une transformation inverse de l'espace de données entrées des pixels vers l'espace des abondances. L'ajout des informations spatiales sous forme "variation totale" est également introduit pour rendre l'algorithme plus robuste au bruit. Néanmoins, le problème d'obtention des données de réalité terrain nécessaires pour l'étape d'apprentissage limite l'application de ce type d'algorithmes. / In this thesis , we present several aspects of hyperspectral imaging technology , while focusing on the problem of non- linear unmixing . We have proposed three solutions for this task. The first one is integrating the advantages of manifold learning in classical unmixing methods to design their nonlinear versions . Results with data generated on a well-known manifold- the " Swissroll " - seem promising. The methods work much better with the increase in non- linearity compared with their linear version. However, the absence of constraint of non- negativity in these methods remains an open question for improvements . The second proposal is using the pre-image method for estimating an inverse transformation of the data form pixel space to abundance of space . The adoption of spatial information as " total variation " is also introduced to make the algorithm more robust to noise . However, the problem of obtaining ground truth data required for learning step limits the application of such algorithms.

No description available.