En raison de la nature cohérente du signal RADAR à synthèse d’ouverture (RSO), les images RSO polarimétriques (RSOPOL) sont affectées par le bruit de chatoiement. L’effet du chatoiement peut être sévère au point de rendre inutilisable la donnée RSOPOL. Ceci est particulièrement vrai pour les données à une vue qui souffrent d’un chatoiement très intense.Un filtrage du bruit est nécessaire pour améliorer l’estimation des paramètres polarimétriques pouvant être calculés à partir de ce type de données. Cette opération constitue une étape importante dans le traitement et l’analyse des images RSOPOL. Récemment une nouvelle approche est apparue en traitement de données visant la solution d’une multitude de problèmes dont le filtrage, la restauration d’images, la reconnaissance de la parole, la classification ou la segmentation d’images. Cette approche est l’apprentissage profond et les réseaux de neurones à convolution (RNC). Des travaux récents montrent que les RNC sont une alternative prometteuse pour le filtrages des images RSO. En effet par leur capacité d’apprendre un modèle optimal de filtrage, ils tendent à surpasser les approches classiques du filtrage sur les images RSO. L’objectif de cette présente étude est d’analyser et d’évaluer l’efficacité du filtrage par RNC sur des données RSOPOL simulées et sur des images satellitaires RSOPOL RADARSAT-2, ALOS/PalSAR et GaoFen-3 acquises sur la région urbaine de San Francisco (Californie). Des modèles inspirés de l’architecture d’un RNC utilisé notamment en Super-résolution ont été adaptés pour le filtrage de la matrice de cohérence polarimétrique. L’effet de différents paramètres structuraux de l’architecture des RNC sur le filtrage ont été analysés, parmi ceux-ci on retrouve entre autres la profondeur du réseau (le nombre de couches empilées), la largeur du réseau (le nombre de filtres par couches convolutives) et la taille des filtres de la première couche convolutive. L’apprentissage des modèles a été effectué par la rétropropagation du gradient de l’erreur en utilisant 3 ensembles de données qui simulent la polarimétrie une vue des diffuseurs selon les classes de Cloude-Pottier. Le premier ensemble ne comporte que des zones homogènes.Les deux derniers ensembles sont composés de simulations en patchwork dont l’intensité locale est simulée par des images de texture et de cibles ponctuelles ajoutées au patchwork dans le cas du dernier ensemble. Les performances des différents filtres par RNC ont été mesurées par des indicateurs comprenant l’erreur relative sur l’estimation de signatures polarimétriques et des paramètres de décomposition ainsi que des mesures de distorsion sur la récupération des détails importants et sur la conservation des cibles ponctuelles. Les résultats montrent que le filtrage par RNC des données polarimétriques est soit équivalent ou nettement supérieur aux filtres conventionnellement utilisées en polarimétrie.Les résultats des modèles les plus profonds obtiennent les meilleures performances pour tous les indicateurs sur l’ensemble des données homogènes simulées. Dans le cas des données en patchwork, les résultats pour la restauration des détails sont nettement favorables au filtrage par RNC les plus profonds.L’application du filtrage par RNC sur les images satellitaires RADARSAT-2,ALOS/PalSAR ainsi GaoFen-3 montre des résultats comparables ou supérieurs aux filtres conventionnels. Les meilleurs résultats ont été obtenus par le modèle à 5 couches cachées(si on ne compte pas la couche d’entrée et de sortie), avec 8 filtres 3×3 par couche convolutive, sauf pour la couche d’entrée où la taille des filtres étaient de 9×9. Par contre,les données d’apprentissage doivent être bien ajustées à l’étendue des statistiques des images polarimétriques réelles pour obtenir de bon résultats. Ceci est surtout vrai au niveau de la modélisation des cibles ponctuelles dont la restauration semblent plus difficiles. / Due to the coherent nature of the Synthetic Aperture Radar (SAR) signal, polarimetric SAR(POLSAR) images are affected by speckle noise. The effect of speckle can be so severe as to render the POLSAR data unusable. This is especially true for single-look data that suffer from very intense speckle. Noise filtering is necessary to improve the estimation of polarimetric parameters that can be computed from this type of data. This is an important step in the processing and analysis of POLSAR images. Recently, a new approach has emerged in data processing aimed at solving a multi-tude of problems including filtering, image restoration, speech recognition, classification orimage segmentation. This approach is deep learning and convolutional neural networks(CONVNET). Recent works show that CONVNET are a promising alternative for filtering SAR images. Indeed, by their ability to learn an optimal filtering model only from the data, they tend to outperform classical approaches to filtering on SAR images. The objective of this study is to analyze and evaluate the effectiveness of CONVNET filtering on simulated POLSAR data and on RADARSAT-2, ALOS/PalSAR and GaoFen-3 satellite images acquired over the San Francisco urban area (California). Models inspired by the architecture of a CONVNET used in particular in super-resolution have been adapted for the filtering of the polarimetric coherency matrix. The effect of different structural parameters of theCONVNET architecture on filtering were analyzed, among which are the depth of the neural network (the number of stacked layers), the width of the neural network (the number of filters per convoluted layer) and the size of the filters of the first convolution layer. The models were learned by backpropagation of the error gradient using 3 datasets that simulate single-look polarimetry of the scatterers according to Cloude-Pottier classes. The first dataset contains only homogeneous areas. The last two datasets consist of patchwork simulations where local intensity is simulated by texture images and point target are added to the patchwork in the case of the last dataset. The performance of the different filters by CONVNET was measured by indicators including relative error on the estimation of polarimetric signatures and decomposition parameters as well as distortion measurements on the recovery of major details and on the conservation of point targets.The results show that CONVNET filtering of polarimetric data is either equivalent or significantly superior to conventional polarimetric filters. The results of the deepest models obtain the best performance for all indicators over the simulated homogeneous dataset. Inthe case of patchwork dataset, the results for detail restoration are clearly favourable to the deepest CONVNET filtering. The application of CONVNET filtering on RADARSAT-2, ALOS/PalSAR andGaoFen-3 satellite images shows results comparable or superior to conventional filters. The best results were obtained by the 5 hidden layers model (not counting the input and outputlayers), with 8 filters 3×3 per convolutional layer, except for the input layer where the filtersize was 9×9. On the other hand, the training data must be well adjusted to the statistical range of the real polarimetric images to obtain good results. This is especially true when modeling point targets that appear to be more difficult to restore.

A time-domain backprojection processor for airborne synthetic aperture radar (SAR) has been developed at the University of Massachusetts’ Microwave Remote Sensing Lab (MIRSL). The aim of this work is to produce a SAR processor capable of addressing the motion compensation issues faced by frequency-domain processing algorithms, in order to create well focused SAR imagery suitable for interferometry. The time-domain backprojection algorithm inherently compensates for non-linear platform motion, dependent on the availability of accurate measurements of the motion. The implementation must manage the relatively high computational burden of the backprojection algorithm, which is done using modern graphics processing units (GPUs), programmed with NVIDIA’s CUDA language. An implementation of the Non-Equispaced Fast Fourier Transform (NERFFT) is used to enable efficient and accurate range interpolation as a critical step of the processing. The phase of time- domain processed imagery is dif erent than that of frequency-domain imagery, leading to a potentially different approach to interferometry. This general purpose SAR processor is designed to work with a novel, dual-frequency S- and Ka-band radar system developed at MIRSL as well as the UAVSAR instrument developed by NASA’s Jet Propulsion Laboratory. These instruments represent a wide range of SAR system parameters, ensuring the ability of the processor to work with most any airborne SAR. Results are presented from these two systems, showing good performance of the processor itself.

No description available.

No description available.

Detecting water under vegetation is critical to tracking the status of geological ecosystems like wetlands. Researchers use different methods to estimate water presence, avoiding costly on-site measurements. Optical satellite imagery allows the automatic delineation of water using the concept of the Normalised Difference Water Index (NDWI). Still, optical imagery is subject to visibility conditions and cannot detect water under the vegetation, a typical situation for wetlands. Synthetic Aperture Radar (SAR) imagery works under all visibility conditions. It can detect water under vegetation but requires deep network algorithms to segment water presence, and manual annotation work is required to train the deep models. This project uses DEEPAQUA, a cross-modal knowledge distillation method, to eliminate the manual annotation needed to extract water presence from SAR imagery with deep neural networks. In this method, a deep student model (e.g., UNET) is trained to segment water in SAR imagery. The student model uses the NDWI algorithm as the non-parametric, cross-modal teacher. The key prerequisite is that NDWI works on the optical imagery taken from the exact location and simultaneously as the SAR. Three different deep architectures are tested in this project: UNET, SegNet, and UNET++, and the Otsu method is used as the baseline. Experiments on imagery from Swedish wetlands in 2020-2022 show that cross-modal distillation consistently achieved better segmentation performances across architectures than the baseline. Additionally, the UNET family of algorithms performed better than SegNet with a confidence of 95%. The UNET++ model achieved the highest Intersection Over Union (IOU) performance. However, no statistical evidence emerged that UNET++ performs better than UNET, with a confidence of 95%. In conclusion, this project shows that cross-modal knowledge distillation works well across architectures and removes tedious and expensive manual work hours when detecting water from SAR imagery. Further research could evaluate performances on other datasets and student architectures. / Att upptäcka vatten under vegetation är avgörande för att hålla koll på statusen på geologiska ekosystem som våtmarker. Forskare använder olika metoder för att uppskatta vattennärvaro vilket undviker kostsamma mätningar på plats. Optiska satellitbilder tillåter automatisk avgränsning av vatten med hjälp av konceptet Normalised Difference Water Index (NDWI). Optiska bilder fortfarande beroende av siktförhållanden och kan inte upptäcka vatten under vegetationen, en typisk situation för våtmarker. Synthetic Aperture Radar (SAR)-bilder fungerar under alla siktförhållanden. Den kan detektera vatten under vegetation men kräver djupa nätverksalgoritmer för att segmentera vattennärvaro, och manuellt anteckningsarbete krävs för att träna de djupa modellerna. Detta projekt använder DEEPAQUA, en cross-modal kunskapsdestillationsmetod, för att eliminera det manuella annoteringsarbete som behövs för att extrahera vattennärvaro från SAR-bilder med djupa neurala nätverk. I denna metod tränas en djup studentmodell (t.ex. UNET) att segmentera vatten i SAR-bilder semantiskt. Elevmodellen använder NDWI, som fungerar på de optiska bilderna tagna från den exakta platsen och samtidigt som SAR, som den icke-parametriska, cross-modal lärarmodellen. Tre olika djupa arkitekturer testas i detta examensarbete: UNET, SegNet och UNET++, och Otsu-metoden används som baslinje. Experiment på bilder tagna på svenska våtmarker 2020-2022 visar att cross-modal destillation konsekvent uppnådde bättre segmenteringsprestanda över olika arkitekturer jämfört med baslinjen. Dessutom presterade UNET-familjen av algoritmer bättre än SegNet med en konfidens på 95%. UNET++-modellen uppnådde högsta prestanda för Intersection Over Union (IOU). Det framkom dock inga statistiska bevis för att UNET++ presterar bättre än UNET, med en konfidens på 95%. Sammanfattningsvis visar detta projekt att cross-modal kunskapsdestillation fungerar bra över olika arkitekturer och tar bort tidskrävande och kostsamma manuella arbetstimmar vid detektering av vatten från SAR-bilder. Ytterligare forskning skulle kunna utvärdera prestanda på andra datamängder och studentarkitekturer.

The incorporation of Gallium Nitride (GaN) Power Amplifiers (PAs) into future high power aperture radar systems is certain; however, the introduction of this technology into multifunction radar systems will present new challenges to radar engineers. This dissertation describes a broad investigation into amplitude and phase transients produced by GaN PAs when they are excited with multifunction radar waveforms. These transients are the result of self-heating electrothermal memory effects and are manifested as interpulse instabilities that can negatively impact the coherent processing of multiple pulses. A behavioral model based on a Foster network topology has been developed to replicate the measured amplitude and phase transients accurately. This model has been used to develop a digital predistortion technique that successfully mitigates the impact of the transients. The Moving Target Indicator (MTI) Improvement Factor and the Root Mean Square (RMS) Pulse-to-Pulse Stability are used as metrics to assess the impact of the transients on radar system performance and to test the effectiveness of a novel digital predistortion concept. / Ph. D. / The incorporation of Gallium Nitride (GaN) Power Amplifiers (PAs) into future radar systems is certain, and will present new challenges to radar engineers. This dissertation describes a broad investigation into signal transients produced by GaN PAs when they are excited with a wide variety of RF pulsed waveforms. These waveforms are representative of those used by a radar system to conduct multiple functions or missions. The transients are primarily the result of changes in the GaN PA gain due to self-heating, and are manifested as differences in consecutive pulses. These pulse-to-pulse differences negatively affect the ability of a radar system to extract information from a received echo. A behavioral model based on a Foster network topology has been developed to replicate the measured signal transients accurately. This model has been used to develop a digital predistortion technique that successfully counteracts the transients mitigating the impact of the transients. The Moving Target Indicator (MTI) Improvement Factor and the Root Mean Square (RMS) Pulse-to-Pulse Stability are used as performance metrics to quantify the effect of the transients on radar system performance and to test the effectiveness of a novel digital predistortion concept.

Das Radar mit synthetischer Apertur (Synthetic Aperture Radar - SAR) liefert eine quasi-fotographische Abbildung der beleuchteten Bodenoberfläche mit zusätzlichen Informationen, die von der gesendeten und empfangenen Polarisation der Wellen abhängig sind. Eine nützliche Anwendung der polarimetrischen SAR-Daten liegt bei der Klassifizierung der Bodenstruktur anhand der polarimetrischen Streuungseigenschaften. In diesem Zusammenhang beschäftigt sich die vorliegende Arbeit mit der Entwicklung und Untersuchung neuer polarimetrischen Fokussierungsfunktion für die SAR-Datenverarbeitung mit Hilfe der polarimetrischen Rückstreuungseigenschaft, die zu einer alternativen quantitativen Auswertung der polarimerischen SAR-Daten führen kann. Die physikalische Optik Approximation wird für die numerische Berechnung der rückgestreuten elektrischen Felder der kanonischen Ziele unter SAR-Geometrie unter Berücksichtigung der Polarisationslage verwendet. Aus den rückgestreuten elektrischen Felder werden die polarimetrischen Radarrückstreuquerschnitte berechnet. Ein SAR-Simulator wird zur Datenverarbeitung der E-SAR des DLR entwickelt. Der Ansatz des polarimetrischen Radarrückstreuquerschnittes ermöglicht die approximierte numerische Berechnung der Rückstreuungseigenschaften der kanonischen Ziele sowohl im kopolaren als auch im kreuzpolaren Polarisationsbetrieb. Bei der SAR-Datenverarbeitung werden die Rohdatensätze durch die Referenzfunktion eines Punktzieles in der Entfernungsrichtung verarbeitet. Bei der Azimutkompression werden die vier Referenzfunktionen, das heißt die Referenzfunktion eines Punktzieles, die polarimetrische Fokussierungsfunktion einer flachen Platte, die polarimetrische Fokussierungsfunktion eines Zweifach-Reflektors und die polarimetrische Fokussierungsfunktion eines Dreifach-Reflektors, eingesetzt. Die qunatitativen Auswertung der SAR-Daten werden anhand des Pauli-Zerlegungstheorems, der differentiellen Reflektivität und des linearen Depolarisationsverhältnises durchgeführt.

A sustainable society is a society that satisfies its resource requirements without endangering the sustainability of these resources. The mineral endowment on the African continent is estimated to be the first or second largest of world reserves. Therefore, it is recognised that the African continent still heavily depends on mineral exports as a key contributor to the gross domestic product (GDP) of various countries. These mining activities, however, do introduce primary and secondary environmental degradation factors. They attract communities to these mining areas, light and heavy industrial establishments occur, giving rise to artisanal activities. This study focussed on satellite RS products as an aid to a mine’s operations and the monitoring of its environment. Effective operational mine management and control ensures a more sustainable and profitable lifecycle for mines. Satellite based RS holds the potential to observe the mine and its surrounding areas at high temporal intervals, different spectral wavelengths and spatial resolutions. The combination of SAR and optical information creates a spatial platform to observe and measure the mine’s operations and the behaviour of specific land cover and land use classes over time and contributes to a better understanding of the mining activities and their influence on the environment within a specific geographical area. This study will introduce an integrated methodology to collect, process and analyse spatial information over a specific targeted mine. This methodology utilises a medium resolution land cover base map, derived from Landsat 8, to understand the predominant land cover types of the surrounding area. Using very high resolution mono- and stereoscopic satellite imagery provides a finer scale analysis and identifies changes in features at a smaller scale. Combining these technologies with the synthetic aperture radar (SAR) applications for precise measurement of surface subsidence or upliftment becomes a spatial toolbox for mine management. This study examines a combination of satellite remote sensing products guided by a systematic workflow methodology to integrate spatial results as an aid for mining operations and environmental monitoring. Some of the results that can be highlighted is the successful land cover classification using the Landsat 8 satellite. The land cover that dominated the Kolomela mine area was the “SHRUBLAND/GRASS” class with a 94% coverage and “MINE” class of 2.6%. Sishen mine had a similar dominated land cover characteristic with a “SHRUBLAND/GRASS” class of 90% and “MINE” class of 4.8%. The Pléiades time-series classification analysis was done using three scenes each acquired at a different time interval. The Sishen and Kolomela mine showed especially changes from the bare soil class to the asphalt or mine class. The Pléiades stereoscopic analysis provided volumetric change detection over small, medium, large and recessed areas. Both the Sishen and Kolomela mines demonstrated height profile changes in each selected category. The last category of results focused on the SAR technology to measure within millimetre accuracy the subsidence and upliftment behaviour of surface areas over time. The Royal Bafokeng Platinum tailings pond area was measured using 74 TerraSAR-X scenes. The tailings wall area was confirmed as stable with natural subsidence that occurred in its surrounding area due to seasonal changes of the soil during rainy and dry periods. The Chuquicamata mine as a large open pit copper mine area was analysed using 52 TerraSAR-X scenes. The analysis demonstrated significant vertical surface movement over some of the dumping sites. It is the wish of the researcher that this dissertation and future research scholars will continue to contribute in this scientific field. These contributions can only assist the mining sector to continuously improve its mining operations as well as its monitoring of the primary as well as the secondary environmental impacts to ensure improved sustainability for the next generation. / Environmental Sciences / M. Sc. (Environmental Science)

No description available.

Das Radar mit synthetischer Apertur (Synthetic Aperture Radar - SAR) liefert eine quasi-fotographische Abbildung der beleuchteten Bodenoberfläche mit zusätzlichen Informationen, die von der gesendeten und empfangenen Polarisation der Wellen abhängig sind. Eine nützliche Anwendung der polarimetrischen SAR-Daten liegt bei der Klassifizierung der Bodenstruktur anhand der polarimetrischen Streuungseigenschaften. In diesem Zusammenhang beschäftigt sich die vorliegende Arbeit mit der Entwicklung und Untersuchung neuer polarimetrischen Fokussierungsfunktion für die SAR-Datenverarbeitung mit Hilfe der polarimetrischen Rückstreuungseigenschaft, die zu einer alternativen quantitativen Auswertung der polarimerischen SAR-Daten führen kann. Die physikalische Optik Approximation wird für die numerische Berechnung der rückgestreuten elektrischen Felder der kanonischen Ziele unter SAR-Geometrie unter Berücksichtigung der Polarisationslage verwendet. Aus den rückgestreuten elektrischen Felder werden die polarimetrischen Radarrückstreuquerschnitte berechnet. Ein SAR-Simulator wird zur Datenverarbeitung der E-SAR des DLR entwickelt. Der Ansatz des polarimetrischen Radarrückstreuquerschnittes ermöglicht die approximierte numerische Berechnung der Rückstreuungseigenschaften der kanonischen Ziele sowohl im kopolaren als auch im kreuzpolaren Polarisationsbetrieb. Bei der SAR-Datenverarbeitung werden die Rohdatensätze durch die Referenzfunktion eines Punktzieles in der Entfernungsrichtung verarbeitet. Bei der Azimutkompression werden die vier Referenzfunktionen, das heißt die Referenzfunktion eines Punktzieles, die polarimetrische Fokussierungsfunktion einer flachen Platte, die polarimetrische Fokussierungsfunktion eines Zweifach-Reflektors und die polarimetrische Fokussierungsfunktion eines Dreifach-Reflektors, eingesetzt. Die qunatitativen Auswertung der SAR-Daten werden anhand des Pauli-Zerlegungstheorems, der differentiellen Reflektivität und des linearen Depolarisationsverhältnises durchgeführt.