Les prairies représentent une source importante de biodiversité dans les paysages agricoles qu’il est important de surveiller. Les satellites de nouvelle génération tels que Sentinel-2 offrent de nouvelles opportunités pour le suivi des prairies grâce à leurs hautes résolutions spatiale et temporelle combinées. Cependant, le nouveau type de données fourni par ces satellites implique des problèmes liés au big data et à la grande dimension des données en raison du nombre croissant de pixels à traiter et du nombre élevé de variables spectro-temporelles. Cette thèse explore le potentiel des satellites de nouvelle génération pour le suivi de la biodiversité et des facteurs qui influencent la biodiversité dans les prairies semi-naturelles. Des outils adaptés à l’analyse statistique des prairies à partir de séries temporelles d’images satellites (STIS) denses à haute résolution spatiale sont proposés. Tout d’abord, nous montrons que la réponse spectrotemporelle des prairies est caractérisée par sa variabilité au sein des prairies et parmi les prairies. Puis, pour les analyses statistiques, les prairies sont modélisées à l’échelle de l’objet pour être cohérent avec les modèles écologiques qui représentent les prairies à l’échelle de la parcelle. Nous proposons de modéliser la distribution des pixels dans une prairie par une loi gaussienne. A partir de cette modélisation, des mesures de similarité entre deux lois gaussiennes robustes à la grande dimension sont développées pour la classification des prairies en utilisant des STIS denses: High-Dimensional Kullback-Leibler Divergence et -Gaussian Mean Kernel. Cette dernière est plus performante que les méthodes conventionnelles utilisées avec les machines à vecteur de support (SVM) pour la classification du mode de gestion et de l’âge des prairies. Enfin, des indicateurs de biodiversité des prairies issus de STIS denses sont proposés à travers des mesures d’hétérogénéité spectro-temporelle dérivées du clustering non supervisé des prairies. Leur corrélation avec l’indice de Shannon est significative mais faible. Les résultats suggèrent que les variations spectro-temporelles mesurées à partir de STIS à 10 mètres de résolution spatiale et qui couvrent la période où ont lieu les pratiques agricoles sont plus liées à l’intensité des pratiques qu’à la diversité en espèces. Ainsi, bien que les propriétés spatiales et temporelles de Sentinel-2 semblent limitées pour estimer directement la diversité en espèces des prairies, ce satellite devrait permettre le suivi continu des facteurs influençant la biodiversité dans les prairies. Dans cette thèse, nous avons proposé des méthodes qui prennent en compte l’hétérogénéité au sein des prairies et qui permettent l’utilisation de toute l’information spectrale et temporelle fournie par les satellites de nouvelle génération. / Grasslands are a significant source of biodiversity in farmed landscapes that is important to monitor. New generation satellites such as Sentinel-2 offer new opportunities for grassland’s monitoring thanks to their combined high spatial and temporal resolutions. Conversely, the new type of data provided by these sensors involves big data and high dimensional issues because of the increasing number of pixels to process and the large number of spectro-temporal variables. This thesis explores the potential of the new generation satellites to monitor biodiversity and factors that influence biodiversity in semi-natural grasslands. Tools suitable for the statistical analysis of grasslands using dense satellite image time series (SITS) with high spatial resolution are provided. First, we show that the spectro-temporal response of grasslands is characterized by its variability within and among the grasslands. Then, for the statistical analysis, grasslands are modeled at the object level to be consistent with ecological models that represent grasslands at the field scale. We propose to model the distribution of pixels in a grassland by a Gaussian distribution. Following this modeling, similarity measures between two Gaussian distributions robust to the high dimension are developed for the lassification of grasslands using dense SITS: the High-Dimensional Kullback-Leibler Divergence and the -Gaussian Mean Kernel. The latter outperforms conventional methods used with Support Vector Machines for the classification of grasslands according to their management practices and to their age. Finally, indicators of grassland biodiversity issued from dense SITS are proposed through spectro-temporal heterogeneity measures derived from the unsupervised clustering of grasslands. Their correlation with the Shannon index is significant but low. The results suggest that the spectro-temporal variations measured from SITS at a spatial resolution of 10 meters covering the period when the practices occur are more related to the intensity of management practices than to the species diversity. Therefore, although the spatial and spectral properties of Sentinel-2 seem limited to assess the species diversity in grasslands directly, this satellite should make possible the continuous monitoring of factors influencing biodiversity in grasslands. In this thesis, we provided methods that account for the heterogeneity within grasslands and enable the use of all the spectral and temporal information provided by new generation satellites.

O uso de imagens provenientes de sensores remotos, tal como sistemas acoplados em aviões e satélites, é cada vez mais frequente, uma vez que permite o monitoramento continuo e periódico ao longo do tempo por meio de diversas observações de uma mesma região, por vezes ampla ou de difícil acesso. Essa ferramenta tem se mostrado importante e significativa em aplicações como o mapeamento de solo e fronteiras; acompanhamento de áreas de desmatamento, queimadas e de produção agrícola. Para gerar resultados interpretáveis ao usuário final, essas imagens devem ser processadas. Atualmente, o método de classificação por máxima verossimilhança é o mais empregado para classificação de imagens multiespectrais de sensores remotos, entretanto, por se tratar de uma técnica supervisionada, seus resultados dependem extensivamente da qualidade do conjunto de treinamento, utilizado para definir os parâmetros do método. A seleção de um bom conjunto de treinamento é um processo custoso e inviabiliza a automação da classificação para diversas imagens. O método de classificação por máxima verossimilhança é também paramétrico e portanto exitem algumas suposições quanto a distribuição dos dados que devem ser atendidas, caso contrário a aplicação do método pode gerar resultados ruins. Tendo em vista essas desvantagens do método da máxima verossimilhança, este trabalho propõe um novo método para a classificação de imagens multiespectrais provenientes de sensores remotos de forma que o procedimento seja autônomo, veloz e preciso, minimizando dessa forma os possíveis erros humanos inseridos em etapas intermediárias do processo, tal como a definição de conjuntos de treinamento. O método aqui proposto pertence ao conjunto das redes neurais artificiais (RNAs) e é denominado growing neural gas (GNG). Este método baseia-se no aprendizado não supervisionado de padrões \"naturais\" dentro de um conjunto de dados por meio da criação e adaptação de uma rede mínima de neurônios. Os resultados gerados a partir da classificação pela RNA foram comparados com os métodos mais utilizados na literatura atual, sendo eles o método da máxima verossimilhança e o método k-means. A partir da biblioteca espectral ASTER, mantida e criada parcialmente pela NASA, foram realizadas várias repetições do experimento, que consiste em classificar os dados de acordo com as diferentes classes existentes, e para cada uma destas repetições calculou-se uma medida de acurácia, denominada índice kappa, além do tempo de execução de cada método, de forma que suas médias foram comparadas via intervalo de confiança gerados por bootstrap não paramétrico. Também investigou-se como a análise de componentes principais (ACP), técnica utilizada para reduzir a dimensão dos dados e consequentemente o custo computacional, pode influenciar no desempenho dos métodos, tanto em sua qualidade de classificação quanto em relação ao tempo de execução. Os resultados mostram que o método proposto é superior nos dois aspectos estudados, acurácia e tempo de execução, para a maioria dos fatores aplicados. Mostra-se ainda um exemplo de aplicação prática em que uma imagem multiespectral de satélite não satisfaz as pré-suposições estabelecidas para o uso do método da máxima verossimilhança e verifica-se a diferença entre os métodos com relação a qualidade final da imagem classificada. / The use of images from remote sensors, such as coupled systems in airplanes and satellites, are increasingly being used because they allow continuous and periodic surveillance over time through several observations of some particular area, sometimes large or difficult to access. This sort of image has shown an important and meaningful participation in applications such as soil and borders mapping; surveillance of deforestation, forest fires and agricultural production areas. To generate interpretable results to the end user, these images must be processed. Currently, the maximum likelihood classification method is the most used for multispectral image classification of remote sensing, however, because it is a supervised technique, the results depend extensively on the quality of the training set, used to define the parameters of the method. Selecting a good training set is a costly process and prevents the automation of classification for different images. The maximum likelihood classification method is also parametric, and therefore, some assumptions about the distribution of the data must be met, otherwise the application of the method can generate bad results. In view of these disadvantages of the maximum likelihood method, this dissertation proposes a new, autonomous, fast and accurate method for multispectral remote sensing imagery classification thereby minimizing the possible human errors inserted at intermediate stages of the process, such as the definition of training sets. The method proposed here belongs to the set of artificial neural networks (ANN) and is called growing neural gas (GNG). This method is based on unsupervised learning of \"natural\" patterns in a dataset through the creation and adaptation of a minimum network of neurons. The results generated from the classification by ANN were compared with the most commonly used methods in the literature: the maximum likelihood method and the k-means method. From the spectral library Aster, maintained and made in part by NASA, several replications of the experiment were made, which is to classify the data according to different preestablished classes, and a measure of accuracy called kappa index was calculated for each of the replicates, in addition to the execution time of each method, so that their means were compared via confidence interval generated by nonparametric bootstrap. It was additionally investigated how principal component analysis (PCA), technique which reduces dimension of data and consequently the computational cost, can influence the performance of methods, both in its quality rating and runtime. The results show that the proposed method is superior in both aspects studied, accuracy and runtime, for the majority of applied factors. Furthermore, it is shown an example of a practical application in which a multispectral satellite image does not necessarily meet the established assumptions for using the maximum likelihood method, and there is a difference between the methods, regarding to its final classified image quality.

Forest-cover in the tropics is changing rapidly due to indiscriminate removal of timber from many localities. The main focus of the study is to develop an operational tool for monitoring biomass and carbon pool of tropical forest ecosystems. The method was applied to a test site of Bangladesh. The research used Landsat ETM+, Landsat TM and IRS pan images of 2001, 1992 and 1999 respectively. Geometrically corrected Landsat ETM+ imagery was obtained from USGS and adjusted to the field using GPS. Historical images were corrected using image-to-image registration. Atmospheric correction was done by modified dark object subtraction method. Stratified sampling design based on the remote sensing image was applied for assessing the above-ground biomass and carbon content of the study area. Field sampling was done during 2002-2003. Dbh and height of all the trees inside the sample plots were measured. Field measurement was finally converted to carbon content using allometric relations. Three different methods: stratification, regression and k-nearest neighbors were tested for combining remote sensing image information and field-based terrestrial carbon pool. Additional field sampling was conducted during 2003-2004 for testing the accuracy. Finally regression method was selected. The amount of carbon released and sequestrated from the ecosystem was estimated. The application of the developed method would be quite useful for understating the terrestrial carbon dynamics and global climate change.

Knowing where ships are located is a key factor to support safe maritime transports, harbor management as well as preventing accidents and illegal activities at sea. Present international solutions for geopositioning in the maritime domain exist such as the Automatic Identification System (AIS). However, AIS requires the ships to constantly transmit their location. Real time imaginary based on geostationary satellites has recently been proposed to complement the existing AIS system making locating and tracking more robust. This thesis investigated and compared two machine learning image analysis approaches – Faster R-CNN and SSD with FPN – for detection and tracking of ships in satellite images. Faster R-CNN is a two stage model which first proposes regions of interest followed by detection based on the proposals. SSD is a one stage model which directly detects objects with the additional FPN for better detection of objects covering few pixels. The MAritime SATellite Imagery dataset (MASATI) was used for training and evaluation of the candidate models with 5600 images taken from a wide variety of locations. The TensorFlow Object Detection API was used for the implementation of the two models. The results for detection show that Faster R-CNN achieved a 30.3% mean Average Precision (mAP) while SSD with FPN achieved only 0.0005% mAP on the unseen test part of the dataset. This study concluded that Faster R-CNN is a candidate for identifying and tracking ships in satellite images. SSD with FPN seems less suitable for this task. It is also concluded that the amount of training and choice of hyper-parameters impacted the results.

Nous étudions ici l’intérêt des descripteurs locaux pour les images satellites optiques et radar. Ces descripteurs, par leurs invariances et leur représentation compacte, offrent un intérêt pour la comparaison d’images acquises dans des conditions différentes. Facilement applicables aux images optiques, ils offrent des performances limitées sur les images radar, en raison de leur fort bruit multiplicatif. Nous proposons ici un descripteur original pour la comparaison d’images radar. Cet algorithme, appelé SAR-SIFT, repose sur la même structure que l’algorithme SIFT (détection de points-clés et extraction de descripteurs) et offre des performances supérieures pour les images radar. Pour adapter ces étapes au bruit multiplicatif, nous avons développé un opérateur différentiel, le Gradient par Ratio, permettant de calculer une norme et une orientation du gradient robustes à ce type de bruit. Cet opérateur nous a permis de modifier les étapes de l’algorithme SIFT. Nous présentons aussi deux applications pour la télédétection basées sur les descripteurs. En premier, nous estimons une transformation globale entre deux images radar à l’aide de SAR-SIFT. L’estimation est réalisée à l’aide d’un algorithme RANSAC et en utilisant comme points homologues les points-clés mis en correspondance. Enfin nous avons mené une étude prospective sur l’utilisation des descripteurs pour la détection de changements en télédétection. La méthode proposée compare les densités de points-clés mis en correspondance aux densités de points-clés détectés pour mettre en évidence les zones de changement. / We study here the interest of local features for optical and SAR images. These features, because of their invariances and their dense representation, offer a real interest for the comparison of satellite images acquired under different conditions. While it is easy to apply them to optical images, they offer limited performances on SAR images, because of their multiplicative noise. We propose here an original feature for the comparison of SAR images. This algorithm, called SAR-SIFT, relies on the same structure as the SIFT algorithm (detection of keypoints and extraction of features) and offers better performances for SAR images. To adapt these steps to multiplicative noise, we have developed a differential operator, the Gradient by Ratio, allowing to compute a magnitude and an orientation of the gradient robust to this type of noise. This operator allows us to modify the steps of the SIFT algorithm. We present also two applications for remote sensing based on local features. First, we estimate a global transformation between two SAR images with help of SAR-SIFT. The estimation is realized with help of a RANSAC algorithm and by using the matched keypoints as tie points. Finally, we have led a prospective study on the use of local features for change detection in remote sensing. The proposed method consists in comparing the densities of matched keypoints to the densities of detected keypoints, in order to point out changed areas.

Die vorliegende Bachelorarbeit ergänzt das bereits existierende nicht-photorealistische 3D-Landschaftsmodell im 'Ethno-Nature Park Uch-Enmek' nach Osten. Zentrum der durchgeführten Modellierungsarbeiten im zwei- und dreidimensionalen Raum ist die Siedlung Karakol. Der existente Workflow - von den Primärdaten bis zum 3D-Modell - wird unabhängig getestet und auf Verbesserungsmöglichkeiten untersucht. Die Schwerpunkte liegen dabei auf der Eignung der existenten Quellen für die Modellierung, der Eignung bisher geschaffener Modellierungswerkzeuge sowie der Quantifizierung des Erfassungsaufwands.

Satellite Image Time Series (SITS) are becoming available at high spatial, spectral and temporal resolutions across the globe by the latest remote sensing sensors. These series of images can be highly valuable when exploited by classification systems to produce frequently updated and accurate land cover maps. The richness of spectral, spatial and temporal features in SITS is a promising source of data for developing better classification algorithms. However, machine learning methods such as Random Forests (RFs), despite their fruitful application to SITS to produce land cover maps, are structurally unable to properly handle intertwined spatial, spectral and temporal dynamics without breaking the structure of the data. Therefore, the present work proposes a comparative study of various deep learning algorithms from the Convolutional Neural Network (CNN) family and evaluate their performance on SITS classification. They are compared to the processing chain coined iota2, developed by the CESBIO and based on a RF model. Experiments are carried out in an operational context using with sparse annotations from 290 labeled polygons. Less than 80 000 pixel time series belonging to 8 land cover classes from a year of Sentinel- 2 monthly syntheses are used. Results show on a test set of 131 polygons that CNNs using 3D convolutions in space and time are more accurate than 1D temporal, stacked 2D and RF approaches. Best-performing models are CNNs using spatio-temporal features, namely 3D-CNN, 2D-CNN and SpatioTempCNN, a two-stream model using both 1D and 3D convolutions. / Tidsserier av satellitbilder (SITS) blir tillgängliga med hög rumslig, spektral och tidsmässig upplösning över hela världen med hjälp av de senaste fjärranalyssensorerna. Dessa bildserier kan vara mycket värdefulla när de utnyttjas av klassificeringssystem för att ta fram ofta uppdaterade och exakta kartor över marktäcken. Den stora mängden spektrala, rumsliga och tidsmässiga egenskaper i SITS är en lovande datakälla för utveckling av bättre algoritmer. Metoder för maskininlärning som Random Forests (RF), trots att de har tillämpats på SITS för att ta fram kartor över landtäckning, är strukturellt sett oförmögna att hantera den sammanflätade rumsliga, spektrala och temporala dynamiken utan att bryta sönder datastrukturen. I detta arbete föreslås därför en jämförande studie av olika algoritmer från Konvolutionellt Neuralt Nätverk (CNN) -familjen och en utvärdering av deras prestanda för SITS-klassificering. De jämförs med behandlingskedjan iota2, som utvecklats av CESBIO och bygger på en RF-modell. Försöken utförs i ett operativt sammanhang med glesa annotationer från 290 märkta polygoner. Mindre än 80 000 pixeltidsserier som tillhör 8 marktäckeklasser från ett års månatliga Sentinel-2-synteser används. Resultaten visar att CNNs som använder 3D-falsningar i tid och rum är mer exakta än 1D temporala, staplade 2D- och RF-metoder. Bäst presterande modeller är CNNs som använder spatiotemporala egenskaper, nämligen 3D-CNN, 2D-CNN och SpatioTempCNN, en modell med två flöden som använder både 1D- och 3D-falsningar.

This paper aims to assess land cover change by high-resolution remote satellite images in Bo Trach district, Quang Binh province. The study used eCognition Developer 9.1 to classify SPOT and Sentinal-2 images of the study area. Objects on the images are characterized by values of Channels, including Red, Green, Blue, NIR, Brightness, NDVI, and RIV. Since then, maps of land cover status in the year of 2005, 2010, and 2017 were created with high accuracy 92.22%, 91.28%, 94.22%, respectively. Overlaid three land cover maps to develop the land cover change maps of two periods 2005-2010 (period 1) and 2010-2017 (period 2). The results show that there is a variation in the area of land cover types, especially agriculture and forest land. Of which, agrarian land increased by 7.7% in period 1 and 9.95% in period 2. Whereas, forest land decreased by 0.6% in period 1 and 1.5% in period 2. / Bài báo này nhằm mục đích đánh giá biến động sử dụng đất bằng viễn thám độ phân giải cao tại huyện Bố Trạch, tỉnh Quảng Bình. Nghiên cứu sử dụng phần mềm eCognition Developer 9.1để phân loại ảnh ảnh SPOT và Sentinal-2 trên địa bàn nghiên cứu. Các đặc trưng của đối tượng trên ảnh được xác định dựa trên giá trị độ sáng các Kênh 1, Kênh 2, Kênh 3, Kênh 4, giá trị độ sáng trung bình (Brightness), chỉ số khác biệt thực vật (NDVI) và tỷ số thực vật (RIV). Từ đó xây dựng được các bản đồ lớp phủ mặt đất các năm 2005, 2010, 2017 với độ chính xác lần lượt là 92,22%, 91,28%, 94.22%. Chồng ghép các bản đồ lớp phủ mặt đất, xây dựng bản đồ biến động sử dụng đất giữa hai thời kỳ 2005-2010 và 2010-2017. Kết quả nghiên cứu cho thấy có sự thay đổi giữa các loại hình lớp phủ gồm: đất nông nghiệp tăng khoảng 7,7% giai đoạn 1 và 9,95% giai đoạn 2. Đất lâm nghiệp giảm khoảng 0,6% giai đoạn 1 và 1,5% giai đoạn 2.

New satellite missions have provided High Resolution (HR) Satellite Image Time Series (SITS), offering detailed spatial, spectral, and temporal information for effective monitoring of diverse Earth features including weather, landforms, oceans, vegetation, and agricultural practices. SITS can be used for an accurate understanding of the Land Cover (LC) behavior and providing the possibility of precise mapping of LCs. Moreover, HR SITS presents an unprecedented possibility for the creation and modification of HR Land Cover Change (LCC) and Land Cover Transition (LCT) maps. For the long-term scale, spanning multiple years, it becomes feasible to analyze LCC and the LCTs occurring between consecutive years. Existing methods in literature often analyze bi-temporal images and miss the valuable multi-temporal/multi-annual information of SITS that is crucial for an accurate SITS analysis. As a result, HR SITS necessitates a paradigm shift in processing and methodology development, introducing new challenges in data handling. Yet, the creation of techniques that can effectively manage the high spatial correlation and complementary temporal resolutions of pixels remains paramount. Moreover, the temporal availability of HR data across historical and current archives varies significantly, creating the need for an effective preprocessing to account for factors like atmospheric and radiometric conditions that can affect image reflectance and their applicability in SITS analysis. Flexible and automatic SITS analysis methods can be developed by paying special attention to handling big amounts of data and modeling the correlation and characterization of SITS in space and time. Novel methods should deal with data preparation and pre-processing at large-scale from end-to-end by introducing a set of steps that guarantee reliable SITS analysis while upholding the computational efficiency for a feasible SITS analysis. In this context, the recent strides in deep learning-based frameworks have demonstrated their potential across various image processing tasks, and thus the high relevance for addressing SITS analysis. Deep learning-based methods can be supervised or unsupervised considering their learning process. Supervised deep learning methods rely on labeled training data, which can be impractical for large-scale multi-temporal datasets, due to the challenges of manual labeling. In contrast, unsupervised deep learning methods are favored as they can automatically discover temporal patterns and changes without the need for labeled samples, thereby reducing the computational load, making them more suitable for handling extensive SITS. In this scenario, the objectives of this thesis are mainly three. Firstly, it seeks to establish a robust and reliable framework for the precise mapping of LCs by designing novel techniques for time series analysis. Secondly, it aims to utilize the capacities of unsupervised deep learning methods, such as pretrained Convolutional Neural Networks (CNNs), to construct a comprehensive methodology for Change Detection (CD), thereby mitigating complexity and reducing computational requirements in comparison with supervised methods. This involves the efficient extraction of spatial, spectral, and temporal features from complex multi-temporal, multi-spectral SITS. Lastly, the thesis endeavors to develop novel methods for analyzing LCCs occurring over extended time periods, spanning multiple years. This multifaceted approach encompasses the detection of changes, timing identification, and classification of the specific types of LCTs. The efficacy of the innovative methodologies and associated techniques is showcased through a series of experiments conducted on HR SITS datasets, including those from Sentinel-2 and Landsat. These experiments reveal significant enhancements when compared to existing methods that represent the current state-of-the-art.

With the advancement of technology and the development of more sophisticated remote sensing sensor systems, the use of satellite imagery has opened up various fields of exploration and application. There has been an increased interest in analysis of multi-temporal satellite image in the past few years because of the wide variety of possible applications of in both short-term and long-term image analysis. The type of changes that might be of interest can range from short-term phenomena such as flood assessment and crop growth stage, to long-term phenomena such as urban fringe development. This thesis studies flood assessment and land cover mapping of satellite images, and proposes nature inspired algorithms that can be easily implemented in realistic scenarios. Disaster monitoring using space technology is one of the key areas of research with vast potential; particularly flood based disasters are more challenging. Every year floods occur in many regions of the world and cause great losses. In order to monitor and assess such situations, decision-makers need accurate near real-time knowledge of the field situation. How to provide actual information to decision-makers for effective flood monitoring and mitigation is an important task, from the point of view of public welfare. Over-estimation of the flooded area leads to over-compensation to people, while under-estimation results in production loss and negative impacts on the population. Hence it is essential to assess the flood damage accurately, both in qualitative and quantitative terms. In such situations, land cover maps play a very critical role. Updating land cover maps is a time consuming and costlier operation when it is performed using traditional or manual methods. Hence, there is a need to find solutions for such problem through automation. Design of automatic systems dedicated to satellite image processing which involves change detection to discriminate areas of land cover change between imaging dates. The system integrates the spectral and spatial information with the techniques of image registration and pattern classification using nature inspired techniques. In the literature, various works have been carried out for solving the problem of image registration and pattern classification using conventional methods. Many researchers have proved, for different situations, that nature inspired techniques are promising in comparison with that of conventional methods. The main advantage of nature inspired technique over any other conventional methods is its stochastic nature, which converges to optimal solution for any dynamic variation in a given satellite image. Results are given in such terms as to delineate change in multi-date imagery using change-versus-no-change information to guide multi-date data analysis. The main objective of this study is to analyze spatio-temporal satellite data to bring out significant changes in the land cover map through automated image processing methods. In this study, for satellite image analysis of flood assessment and land cover mapping, the study areas and images considered are: Multi-temporal MODerate-resolution Imaging Spectroradiometer (MODIS) image around Krishna river basin in Andhra Pradesh India; Linear Imaging Self Scanning Sensor III (LISS III)and Synthetic Aperture Radar(SAR)image around Kosi river basin in Bihar, India; Landsat7thematicmapperimage from the southern part of India; Quick-Bird image of the central Bangalore, India; Hyperion image around Meerut city, Uttar Pradesh, India; and Indian pines hyperspectral image. In order to develop a flood assessment framework for this study, a database was created from remotely sensed images (optical and/or Synthetic Aperture Radar data), covering a period of time. The nature inspired techniques are used to find solutions to problems of image registration and pattern classification of a multi-sensor and multi-temporal satellite image. Results obtained are used to localize and estimate accurately the flood extent and also to identify the type of the inundated area based on land cover mapping. The nature inspired techniques used for satellite image processing are Artificial Neural Network (ANN), Genetic Algorithm (GA),Particle Swarm Optimization (PSO), Firefly Algorithm(FA),Glowworm Swarm Optimization(GSO)and Artificial Immune System (AIS). From the obtained results, we evaluate the performance of the methods used for image registration and pattern classification to compare the accuracy of satellite image processing using nature inspired techniques. In summary, the main contributions of this thesis include (a) analysis of flood assessment and land cover mapping using satellite images and (b) efficient image registration and pattern classification using nature inspired algorithms, which are more popular than conventional optimization methods because of their simplicity, parallelism and convergence of the population towards the optimal solution in a given search space.