Advanced Kernel Methods For Remote Sensing Image Classification

Résumé Suite aux recentes avancées technologiques, les archives d'images digitales ont connu une croissance qualitative et quantitative sans précédent. Malgré les énormes possibilités qu'elles offrent, ces avancées posent de nouvelles questions quant au traitement des masses de données saisies. Cette question est à la base de cette Thèse: les problèmes de traitement d'information digitale à très haute résolution spatiale et/ou spectrale y sont considérés en recourant à des approches d'apprentissage statistique, les méthodes à noyau. Cette Thèse étudie des problèmes de classification d'images, c'est à dire de catégorisation de pixels en un nombre réduit de classes refletant les propriétés spectrales et contextuelles des objets qu'elles représentent. L'accent est mis sur l'efficience des algorithmes, ainsi que sur leur simplicité, de manière à augmenter leur potentiel d'implementation pour les utilisateurs. De plus, le défi de cette Thèse est de rester proche des problèmes concrets des utilisateurs d'images satellite sans pour autant perdre de vue l'intéret des méthodes proposées pour le milieu du machine learning dont elles sont issues. En ce sens, ce travail joue la carte de la transdisciplinarité en maintenant un lien fort entre les deux sciences dans tous les développements proposés. Quatre modèles sont proposés: le premier répond au problème de la haute dimensionalité et de la redondance des données par un modèle optimisant les performances en classification en s'adaptant aux particularités de l'image. Ceci est rendu possible par un système de ranking des variables (les bandes) qui est optimisé en même temps que le modèle de base: ce faisant, seules les variables importantes pour résoudre le problème sont utilisées par le classifieur. Le manque d'information étiquétée et l'incertitude quant à sa pertinence pour le problème sont à la source des deux modèles suivants, basés respectivement sur l'apprentissage actif et les méthodes semi-supervisées: le premier permet d'améliorer la qualité d'un ensemble d'entraînement par interaction directe entre l'utilisateur et la machine, alors que le deuxième utilise les pixels non étiquetés pour améliorer la description des données disponibles et la robustesse du modèle. Enfin, le dernier modèle proposé considère la question plus théorique de la structure entre les outputs: l'intègration de cette source d'information, jusqu'à présent jamais considérée en télédétection, ouvre des nouveaux défis de recherche. Advanced kernel methods for remote sensing image classification Devis Tuia Institut de Géomatique et d'Analyse du Risque September 2009 Abstract The technical developments in recent years have brought the quantity and quality of digital information to an unprecedented level, as enormous archives of satellite images are available to the users. However, even if these advances open more and more possibilities in the use of digital imagery, they also rise several problems of storage and treatment. The latter is considered in this Thesis: the processing of very high spatial and spectral resolution images is treated with approaches based on data-driven algorithms relying on kernel methods. In particular, the problem of image classification, i.e. the categorization of the image's pixels into a reduced number of classes reflecting spectral and contextual properties, is studied through the different models presented. The accent is put on algorithmic efficiency and the simplicity of the approaches proposed, to avoid too complex models that would not be used by users. The major challenge of the Thesis is to remain close to concrete remote sensing problems, without losing the methodological interest from the machine learning viewpoint: in this sense, this work aims at building a bridge between the machine learning and remote sensing communities and all the models proposed have been developed keeping in mind the need for such a synergy. Four models are proposed: first, an adaptive model learning the relevant image features has been proposed to solve the problem of high dimensionality and collinearity of the image features. This model provides automatically an accurate classifier and a ranking of the relevance of the single features. The scarcity and unreliability of labeled. information were the common root of the second and third models proposed: when confronted to such problems, the user can either construct the labeled set iteratively by direct interaction with the machine or use the unlabeled data to increase robustness and quality of the description of data. Both solutions have been explored resulting into two methodological contributions, based respectively on active learning and semisupervised learning. Finally, the more theoretical issue of structured outputs has been considered in the last model, which, by integrating outputs similarity into a model, opens new challenges and opportunities for remote sensing image processing.

ZHARP: three-dimensional motion tracking from a single image plane.

Three-dimensional imaging and quantification of myocardial function are essential steps in the evaluation of cardiac disease. We propose a tagged magnetic resonance imaging methodology called zHARP that encodes and automatically tracks myocardial displacement in three dimensions. Unlike other motion encoding techniques, zHARP encodes both in-plane and through-plane motion in a single image plane without affecting the acquisition speed. Postprocessing unravels this encoding in order to directly track the 3-D displacement of every point within the image plane throughout an entire image sequence. Experimental results include a phantom validation experiment, which compares zHARP to phase contrast imaging, and an in vivo study of a normal human volunteer. Results demonstrate that the simultaneous extraction of in-plane and through-plane displacements from tagged images is feasible.

Classification of urban multi-angular image sequences by aligning their manifolds

When dealing with multi-angular image sequences, problems of reflectance changes due either to illumination and acquisition geometry, or to interactions with the atmosphere, naturally arise. These phenomena interplay with the scene and lead to a modification of the measured radiance: for example, according to the angle of acquisition, tall objects may be seen from top or from the side and different light scatterings may affect the surfaces. This results in shifts in the acquired radiance, that make the problem of multi-angular classification harder and might lead to catastrophic results, since surfaces with the same reflectance return significantly different signals. In this paper, rather than performing atmospheric or bi-directional reflection distribution function (BRDF) correction, a non-linear manifold learning approach is used to align data structures. This method maximizes the similarity between the different acquisitions by deforming their manifold, thus enhancing the transferability of classification models among the images of the sequence.

Edulcorants: entre mythe et réalité [Sweeteners: between myth and reality]

As the prevalence of obesity and diabetes are continually increasing, the use of "false sugars" otherwise known as sweeteners, and their associated health issues are being more and more discussed. A higher sugared power, less calories as well as a moderated or non-existent effect on blood sugar would lead to believe that sweeteners are helpful. However, we CANNOT say that they are THE solution as they can contain calories, may have some undesired effects, and moreover they ease the conscience without actually allowing a weight loss with their sole use. They are to be used with judgment, wittingly and especially when comparing sweetened products. The sweetener myth is often far from reality. It is therefore important to give our patients the means to analyze their dietary intake with regard to their sweeteners ingestion.

High-grade gliomas: reality and hopes.

In this issue of the Chinese Journal of Cancer, European experts review current standards, trends, and future prospects in the difficult domain of high-grade glioma. In all fields covered by the different authors, the progress has been impressive. For example, discoveries at the molecular level have already impacted imaging, surgery, radiotherapy, and systemic therapies, and they are expected to play an increasing role in the management of these cancers. The European Organization for Research and Treatment of Cancer (EORTC) has pioneered new treatment strategies and contributed to new standards. The articles in this issue will cover basic molecular biological principles applicable today, novel surgical approaches, innovations in radiotherapy planning and delivery, evidence-based standards for radiotherapy alone or combined with chemotherapy, current standards and novel approaches for systemic treatments, and the important but often neglected field of health-related quality of life. Despite the advances described in these articles, the overall prognosis of high-grade glioma, especially glioblastoma, remains poor, and more research is needed to address this problem.

Drug targeting strategies for photodynamic therapy.

In human pathologies, therapeutic treatments are often limited by the lack of selectivity of drugs and their elevated effective concentrations. Targeting these agents to a defined tissue could enhance their selectivity and then diminish their side effects when compared to drugs that accumulate in the entire body. Targeting could also improve treatment efficiency by allowing a localized high concentration of the agents. Based on the different behaviors and patterns of expression between diseased and normal cells, strategies for targeting can be explored. For example, receptors, proteases or trans-membrane carriers could be different or differently expressed. Many therapeutic procedures rely on this fact, including photodynamic therapy (PDT). PDT is already used in the treatment of some cancers, of inflammatory diseases and others diseases such as age-related macular degeneration or acne. PDT relies on the activation of a photosensitizer (PS) by visible light which results in the production of cytotoxic reactive oxygen species. In PDT, the general distribution of PS to the whole body leads to generalized photosensitization and poor acceptance of treatments by patients. One way to avoid these effects is to improve the targeting of PSs to diseased tissues using modification of PS with peptides or proteins that will target specific receptors or enzymes. PSs could also be functionalized with non-proteic ligands such as organometalics to achieve targeted and/or combined therapies. Alternatively, PSs could be encapsulated in nanoparticles bearing targeting agents which will decrease concentration of free circulating PS and improve photodynamic efficiency. These different approaches will be discussed in the present review with an emphasis on the use of peptides and proteins.