Éditeur(s) : IRD : Institut de Recherche pour le Développement Université des Antilles et de la Guyane
Extrait de : "Projet CARIBSAT, système caribéen d'information environnementale" : conférence, le 21 mars 2013. Université des Antilles et de la Guyane
Description : L'intervention débute en pointant l'accentuation de l'érosion des sols sur les bassins versants, à cause de l'occupation du sol et des activités anthropiques. Une partie de ces particules érodées se retrouve à l'exutoire littoral, influant sur la qualité des eaux de mer, et donc sur la santé des récifs coralliens. Il faudrait donc pouvoir modéliser le transit de matière des bassins versants aux zones côtières. Pour ce faire, plusieurs étapes sont prévues : classification de l'occupation du sol par des images satellites SPOT5, développement d'un modèle de sensibilité à l'érosion, élaboration d'indicateurs spatialisés sur les bassins versants et à leur exutoire.
Siècle(s) traité(s) : 21
Droits : CC-BY-NC-ND - Attribution - Pas d'utilisation commerciale - Pas de modification
Permalien : http://www.manioc.org/fichiers/V13092
V13092

Éditeur(s) : IRD : Institut de Recherche pour le Développement Université des Antilles et de la Guyane IRD : Institut de Recherche pour le Développement Université des Antilles et de la Guyane
Extrait de : "Projet CARIBSAT, système caribéen d'information environnementale" : conférence, le 21 mars 2013. Université des Antilles et de la Guyane
Résumé : L'intervention traite de l'usage d'images satellites à l'étude des milieux marins. La cartographie des fonds marins renseigne sur la répartition des habitats et des biocénoses marines, permet un suivi temporel d'évolution et fournit des outils pour la gestion de l'environnement. Les méthodes employées sont la plongée sous-marine, la cartographie acoustique, la photographie aérienne et l'imagerie satellite. Cette dernière est utile à plusieurs titres : fréquence quotidienne des données, meilleure résolution d'images, analyse de séries temporelles, automatisation des méthodes de classification, gain de temps et de réduction de coûts. Des applications techniques concrètes complètent la présentation.
L'intervention traite de l'usage d'images satellites à l'étude des milieux marins. La cartographie des fonds marins renseigne sur la répartition des habitats et des biocénoses marines, permet un suivi temporel d'évolution et fournit des outils pour la gestion de l'environnement. Les méthodes employées sont la plongée sous-marine, la cartographie acoustique, la photographie aérienne et l'imagerie satellite. Cette dernière est utile à plusieurs titres : fréquence quotidienne des données, meilleure résolution d'images, analyse de séries temporelles, automatisation des méthodes de classification, gain de temps et de réduction de coûts. Des applications techniques concrètes complètent la présentation.
Siècle(s) traité(s) : 21
Droits : CC-BY-NC-ND - Attribution - Pas d'utilisation commerciale - Pas de modification
Permalien : http://www.manioc.org/fichiers/V13088
V13088
V13088
V13088

Résumé : In the beginning, the coastal fleet of the Bigouden region fished by trawling and was completely focused on exploiting lobster and hake. It is a traditional business that expanded between the last war and the 70s, reaching its peak between 1975 and 1980. The best image to give an idea of this fleet is that of a "bucket waterwheel" endlessly discharging its catch in the four ports of the Guilvinec district. Where there is a seafaring business in Guilvinec, Loctudy and Saint-Guénolé, there is a complementarity that is expressed in two ways: - coastal fishing constitutes a very noticeable and regular supplement to production, since sales happen daily and only bad winter weather causes a relative decrease in the supplement. - coastal fishing gives a flattering image of the port based on the display of high-quality leading market products: daily live lobster, hake and angler. In this study, we have above all tried to evaluate the situation of the Bigouden coastal system as an entity, while including in the same analysis the species, fleets and the exploitation of the resource. For the main species, a balance sheet is proposed. It covers the lobster, hake and angler separately. Lobster, because of its small amount of movement, can be considered as being an isolated stock, while fish belong to much more wide-ranging groups. The lobster has gained from various increases in meshing that have been imposed since 1971, even if these have not always been strictly applied. One of the results of this management by means of meshing is a "depreciation" of the variations of daily yields around an average value of 100kg. For the hake, the dominant trait is an increase in yields in the last few years in accordance with an improvement of the situation in the Gulf of Gascony due to a generalised decrease of effort and a modification of the exploitation profile since the appearance of boulters and straight nets. Whereas the situation for lobster fisheries, as regards the hake's marketable size, did not improve, since the demand for small hake still exists, we have nonetheless been witnessing for about ten years a slow diversification of the fleets' business activities, which is much more in line with sequential hake fisheries' logical outline of exploitation. When lobster yields decrease, one part of the fleet focuses on the fish at the edges of the lobster fisheries. For the angler, we see rather similar phenomena and a significant increase in unloading of the two species by lobster fishermen due also to a widening of their fishing areas. On the lobster mudflats themselves, there is no notable change in exploitation in the historical records. All studies of coastal fishing always come up against the problem of acquiring statistical data and the Bigouden region is no exception to the rule. To improve this situation, in 1987 and 1988 two items were chosen: - a better awareness of the activity and of the areas frequented, by survey during unloading in Guilvinec and in Loctudy. - a detailed description of the haul by species, by sampling, in Loctudy, on two-thirds of coastal sales, between December 1987 and November 1988. For this reason, analysis of fleets by type is still cursory and covers only one year. The appearance of new trends towards fish can only be described qualitatively, since there is no historical record. The classification of boats was done by analysing data furnished by the auctions. This first assessment is not necessarily reliable, above all because of the auctions' inaccurate description of the hauls, a deficiency that we have tried to overcome by surveying in Loctudy. In conclusion, we are witnessing in the Bigouden region, as in other entities, a rebalancing between traditional business activities and modern trends. Lobster fishing, almost exclusive before the 1980s, is tending to decrease for two main reasons: - a generalised decrease in activity in the lobster fisheries due to non-renewal of coastal fish fleets, - The transfer of part of the lobster fishermen's activities to fish, since the lobster and its accompanying catch no longer seemed able to insure the economic viability of new ships. Other maritime districts resolved their lobster problem in a different way. The Lorient fishermen improved their yields by using more capable trawlers, the fishermen of Sable by exploiting the hake fish run to the fullest and limiting their lobster fishing to the best season. The Bigouden region is slowly turning to the second solution, as it renews its fleet.
A ses origines, la flottille côtière du pays bigouden péchant au chalut, est tout entière orientée vers l'exploitation de la langoustine et du merlu. Il s'agit d'une activité traditionnelle qui s' amplifie entre la dernière guerre et les années 70 pour atteindre son apogée entre 1975 et 1980. La meilleure image que l'on puisse donner de cette flottille est celle d'une "noria" débarquant inlassablement sa capture dans les quatre ports du quartier du Guilvinec. Là où existe une activité hauturière à Guilvinec, Loctudy et Saint-Guénolé, il y a une complémentarité qui s'exprime de deux manières : - la pêche côtière constitue un appoint très appréciable et régulier de production, puisque la vente est quotidienne et que seules les intempéries hivernales entraînent une baisse relative des apports. - la pêche côtière donne du port une image flatteuse basée sur la présentation de produits d'appel de grande qualité : langoustine vivante, merlu et baudroies quotidiens. Dans cette étude, nous avons surtout tenté de faire le point sur le système côtier bigouden en tant qu'entité, en englobant dans la même analyse des espèces, les flottilles et l'exploitation de la ressource. Pour les espèces principales, un bilan est proposé. Il couvre séparément la langoustine, le merlu et les baudroies. La langoustine, en raison de ses faibles déplacements, peut être considérée comme constituant un stock isolé, alors que les poissons appartiennent à des ensembles beaucoup plus vastes. La langoustine a beaucoup gagné des diverses augmentations de maillage qui ont été imposées depuis 1971, même si elles n'ont pas toujours été appliquées avec rigueur. Un des résultats de cette gestion par les maillages est un "amortissement" des variations des rendements journaliers autour d'une valeur moyenne de 100 kg. Pour le merlu, le trait dominant est une augmentation des rendements dans les dernières années en accord avec une amélioration de la situation dans le Golfe de Gascogne due à une baisse généralisée de l'effort et à une modification du profil d'exploitation depuis l'apparition des palangriers et des filets droits. Si la situation sur les pêcheries de langoustine, au regard de la taille marchande du merlu, ne s'est pas améliorée puisque la demande de petits merluchons existe toujours, on assiste par contre depuis une dizaine d'années à une diversification lente des activités des flottilles entrant beaucoup mieux dans le schéma logique d'exploitation des pêcheries séquentielles de merlu. Quand les rendements en langoustine baissent, une partie de la flottille s'oriente vers les poisson sur les marges des pêcheries de langoustine. Pour les baudroies on observe des phénomènes assez analogues et une augmentation importante des débarquements des deux espèces par les langoustiniers due également à un élargissement de leurs zones de pêche. Sur les vasières à langoustine elles-mêmes, il n'y a pas de modification notable de l'exploitation dans la série historique. Toute étude des pêches côtières se heurte toujours au problème d'acquisition des données statistiques et le pays bigouden n'échappe pas à la règle. Pour améliorer cette situation, en 1987 et 1988, deux points ont été retenus : - une meilleure connaissance de l'effort et des zones fréquentées, par enquête au moment des débarquements à Guilvinec et à Loctudy, - une description détaillée des apports par espèce, par échantillonnage, à Loctudy, sur les 2/3 des ventes côtières, entre décembre 1987 et novembre 1988. Pour cette raison, l'analyse de flottilles en types est encore sommaire et n'a porté que sur une année. L'apparition des tendances nouvelles vers le poisson ne peut être décrite que qualitativement puisque la série historique n'existe pas. La classification des bateaux a été faite par l'analyse des données fournies par les criées. Ce premier diagnostic ne présente pas toutes les garanties en raison surtout d'une description peu fidèle des apports par les criées, déficience que l'on a essayé de pallier en enquêtant à Loctudy. En conclusion, on assiste dans le pays bigouden comme dans d'autres entités à un rééquilibrage entre les activités traditionnelles et des orientations modernes. La pêche de langoustine, quasi exclusive avant les années 1980, tend à décroître pour deux raisons principales : - une baisse généralisée de l'effort sur les pêcheries de langoustine due au nonrenouvellement des flottilles de pêche côtière, - le report d'une partie de l'effort des langoustiniers vers les pêcheries de poissons, la langoustine et sa capture accessoire ne semblant plus pouvoir assurer la rentabilité économique des navires neufs. D'autres quartiers maritimes ont résolu de façon différente leur problème langoustine. Les Lorientais ont amélioré leurs rendements en adoptant des chaluts plus performants, les Sablais en exploitant à fond la séquence des pêcheries de merlu et en limitant leur intervention sur la langoustine à la saison la plus favorable. Le pays bigouden s'engage lentement dans la seconde solution, au rythme du renouvellement de sa flottille.

Droits : info:eu-repo/semantics/openAccess
http://archimer.ifremer.fr/doc/1989/rapport-2202.pdf
http://archimer.ifremer.fr/doc/00000/2202/

Éditeur(s) : HAL CCSD Springer Verlag
Résumé : International audience
Pituitary adenomas are currently classified by histological, immunocytochemical and numerous ultrastructural characteristics lacking unequivocal prognostic correlations. We investigated the prognostic value of a new clinicopathological classification with grades based on invasion and proliferation. This retrospective multicentric case-control study comprised 410 patients who had surgery for a pituitary tumour with long-term follow-up. Using pituitary magnetic resonance imaging for diagnosis of cavernous or sphenoid sinus invasion, immunocytochemistry, markers of the cell cycle (Ki-67, mitoses) and p53, tumours were classified according to size (micro, macro and giant), type (PRL, GH, FSH/LH, ACTH and TSH) and grade (grade 1a: non-invasive, 1b: non-invasive and proliferative, 2a: invasive, 2b: invasive and proliferative, and 3: metastatic). The association between patient status at 8-year follow-up and age, sex, and classification was evaluated by two multivariate analyses assessing disease- or recurrence/progression-free status. At 8 years after surgery, 195 patients were disease-free (controls) and 215 patients were not (cases). In 125 of the cases the tumours had recurred or progressed. Analyses of disease-free and recurrence/progression-free status revealed the significant prognostic value (p < 0.001; p < 0.05) of age, tumour type, and grade across all tumour types and for each tumour type. Invasive and proliferative tumours (grade 2b) had a poor prognosis with an increased probability of tumour persistence or progression of 25- or 12-fold, respectively, as compared to non-invasive tumours (grade 1a). This new, easy to use clinicopathological classification of pituitary endocrine tumours has demonstrated its prognostic worth by strongly predicting the probability of post-operative complete remission or tumour progression and so could help clinicians choose the best post-operative therapy.
ISSN: 0001-6322

hal-00875723
https://hal.archives-ouvertes.fr/hal-00875723
DOI : 10.1007/s00401-013-1084-y
PUBMED : 23400299

Éditeur(s) : HAL CCSD
Résumé : International audience
This paper shows the information criteria (IC) performances in feature selection framework. Feature selection aims to select a representative subset among a wide set of features. We apply this approach to classify an hand segmented image. The performance is tested using various feature selection schemes (SFS, SBS, SFFS and SBFS) to select the candidate subsets. The accuracy of the approach is based on a good quality of the joint probability density approximation of the combined features. They are obtained using histogram optimized thanks to the adaptive arithmetic coding principles. Our approach is tested on different reference data. The subsets quality is evaluated using correct classification rate computed on multiple classifiers. Results show stability and convergence properties of this tool and its ability to select representative subsets (in the sense that the subset of feature is a good characterization of the classes in which the data belong). Information Criteria could be used for feature selection as a good alternative to other criteria.
CISP Proceedings
Shangay, China

hal-00634749
https://hal.archives-ouvertes.fr/hal-00634749

Éditeur(s) : HAL CCSD IEEE
Résumé : International audience
This paper proposes a novel automated approach for the categorization of cells in fluorescence microscopy images. Our supervised classification method aims at recognizing patterns of unlabeled cells based on an annotated dataset. First, the cell images need to be indexed by encoding them in a feature space. For this purpose, we propose tailored bio-inspired features relying on the distribution of contrast information. Then, a supervised learning algorithm is proposed for classifying the cells. We carried out experiments on cellular images related to the diagnosis of autoimmune diseases, testing our classification method on the HEp-2 Cells dataset of Foggia et al (CBMS 2010). Results show classification precision larger than 96% on average, thus confirming promising application of our approach to the challenging application of cellular image classification for computer-aided diagnosis.
ICPR 2012
Tsukuba, Japan

hal-00958856
https://hal.archives-ouvertes.fr/hal-00958856
https://hal.archives-ouvertes.fr/hal-00958856/document
https://hal.archives-ouvertes.fr/hal-00958856/file/icpr_2012.pdf

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Voting rules relying on k-nearest neighbors (k-NN) are an effective tool in countless many machine learning techniques. Thanks to its simplicity, k-NN classification is very attractive to practitioners, as it enables very good performances in several practical applications. However, it suffers from various drawbacks, like sensitivity to "noisy" instances and poor generalization properties when dealing with sparse high-dimensional data. In this paper, we tackle the k-NN classification problem at its core by providing a novel k-NN boosting approach. Namely, we propose a supervised learning algorithm, called Universal Nearest Neighbors (UNN), that induces a leveraged k-NN rule by globally minimizing a surrogate risk upper bounding the empirical misclassification rate over training data. Interestingly, this surrogate risk can be arbitrary chosen from a class of Bregman loss functions, including the familiar exponential, logistic and squared losses. Furthermore, we show that UNN allows to efficiently filter a dataset of instances by keeping only a small fraction of data. Experimental results on the synthetic Ripley's dataset show that such a filtering strategy is able to reject "noisy" examples, and yields a classification error close to the optimal Bayes error. Experiments on standard UCI datasets show significant improvements over the current state of the art.
ISSN: 0925-2312

hal-00664462
https://hal.inria.fr/hal-00664462

Éditeur(s) : HAL CCSD
Résumé : International audience
In this paper we propose a hybrid approach using mutual information and multi-objective optimization for feature subset selection problem. The hybrid aspect is due to the sequence of a filter method and a wrapper method in order to take advantages of both. The filter method reduces the exploration space by keeping subsets having good internal properties and the wrapper method chooses among the remaining subsets with a classification performances criterion. In the filter step, the subsets are evaluated in a multi-objective way to ensure diversity within the subsets. The evaluation is based on the mutual information to estimate the dependency between features and classes and the redundancy between features within the same subset. We kept the non-dominated (Pareto optimal) subsets for the second step. In the wrapper step, the selection is made according to the stability of the subsets regarding classification performances during learning stage on a set of classifiers to avoid the specialization of the selected subsets for a given classifiers. The proposed hybrid approach is experimented on a variety of reference data sets and compared to the classical feature selection methods FSDD and mRMR. The resulting algorithm outperforms these algorithms and the computation complexity remains acceptable even if it increases with regards to these two fast selection methods.
Journal of Informatics and Mathematical Sciences

hal-01375982
https://hal.archives-ouvertes.fr/hal-01375982

Éditeur(s) : HAL CCSD
Résumé : International audience
The k-nearest neighbors (k-NN) classification rule is still an essential tool for computer vision applications, such as scene recognition. However, k-NN still features some major drawbacks, which mainly reside in the uniform voting among the nearest prototypes in the feature space. In this paper, we propose a new method that is able to learn the "relevance" of prototypes, thus classifying test data using a weighted k-NN rule. In particular, our algorithm, called Multi-class Leveraged k-nearest neighbor (MLNN), learns the prototype weights in a boosting framework, by minimizing a surrogate exponential risk over training data. We propose two main contributions for improving computational speed and accuracy. On the one hand, we implement learning in an inherently multiclass way, thus providing significant computation time reduction over one-versus-all approaches. Furthermore, the leveraging weights enable effective data selection, thus reducing the cost of k-NN search at classification time. On the other hand, we propose a kernel generalization of our approach to take into account real-valued similarities between data in the feature space, thus enabling more accurate estimation of the local class density. We tested MLNN on three datasets of natural images. Results show that MLNN significantly outperforms classic k-NN and weighted k-NN voting. Furthermore, using an adaptive Gaussian kernel provides significant performance improvement. Finally, the best results are obtained when using MLNN with an appropriate learned metric distance.
Proceedings of the 10th Asian Conference on Computer Vision, ACCV 2010, November 8-12, 2010, Queenstown, New Zealand
Queenstown, New Zealand

hal-00664606
https://hal.inria.fr/hal-00664606
https://hal.inria.fr/hal-00664606/document
https://hal.inria.fr/hal-00664606/file/mlnn_accvfinal.pdf

Éditeur(s) : HAL CCSD
Résumé : International audience
The segmentation of digital images of plants is a tricky operation. In the example of a plant image on a unhomogeneous background, i.e. taken in its environment, the colorimetric diversity of the elements of a scene or the large number of forms can amplify the phenomena of over-segmentation. Global segmentation methods such as Mean Shift are then in this case the ones which will give the best results. These methods take into account the totality of the pixels of an image before classifying a point. On the other hand, complexity is increased, because it is necessary to go through the whole image treated, in order to find the mode of the point which one wishes to classify. In this article, we plan to couple the global segmentation with a local method which would take over in the event of obvious classification of a given point. The Mean Shift method is used for this purpose in association with Markov's chains.
Applied Imagery Pattern Recognition (AIPR)
Washington DC, United States

hal-00637143
https://hal.archives-ouvertes.fr/hal-00637143
DOI : 10.1109/AIPR.2011.6176338

Éditeur(s) : HAL CCSD Springer
Résumé : Third International Workshop, MLMI 2012, Held in Conjunction with MICCAI 2012
International audience
Universal Nearest Neighbours (unn) is a classifier recently proposed, which can also effectively estimates the posterior probability of each classification act. This algorithm, intrinsically binary, requires the use of a decomposition method to cope with multiclass problems, thus reducing their complexity in less complex binary subtasks. Then, a reconstruction rule provides the final classification. In this paper we show that the application of unn algorithm in conjunction with a reconstruction rule based on the posterior probabilities provides a classification scheme robust among different biomedical image datasets. To this aim, we compare unn performance with those achieved by Support Vector Machine with two different kernels and by a k Nearest Neighbours classifier, and applying two different reconstruction rules for each of the aforementioned classification paradigms. The results on one private and five public biomedical datasets show satisfactory performance.
Machine Learning in Medical Imaging
Nice, France

hal-00958909
https://hal.archives-ouvertes.fr/hal-00958909
https://hal.archives-ouvertes.fr/hal-00958909/document
https://hal.archives-ouvertes.fr/hal-00958909/file/Miccai.pdf
DOI : 10.1007/978-3-642-35428-1_15

Éditeur(s) : HAL CCSD
Résumé : Co-encadrement assuré par le Groupement pour le Développement de la Télédétection Aérospatiale (GDTA) Diplôme : DESS
il s'agit d'un type de produit dont les métadonnées ne correspondent pas aux métadonnées attendues dans les autres types de produit : DISSERTATION
This document proposes a new map of the Eastern Guyana Shield, with the SPOT4-VEGETATION remotesensed datas of the year 2000. A new synthesis method, with the mean value, was applied and compared with the Maximum Value Method (WVC). To prepare this datas for the synthesis, a cloud filter was defined :BSC (BEST Sky Clear). With this results, an ISODATA classification was proceeded, and, produces 25 classes to detail the tropical maist forest with the 1 km low resolution. To valid this result, this document propose a method with the higthresolution datas Landsat 7 ETM+. This method aims to guide the ground truth and goes thoroughly into the description of the moist tropical forest classes. The sample, used to try this method, shows the boundaries and the constraints of this method. This points are debating in the discussion. Finally, the perspectives to develop the method are presented in the conclusion.
Ce document propose une nouvelle cartographie du Bouclier Oriental Guyanais à l'aide d'images satellites SPOT4-VEGETATION de l'année 2000. Une nouvelle méthode de synthèse annuelle, utilisant la valeur moyenne de la radiométrie, a été appliquée et comparée à la méthode du Maximum Value Composite (MVC). A cette occasion, un filtre de nuage a été défini pour préparer les données à cette synthèse : le filtre BSC (Best Sky Clear). Sur la base de ce résultat, une classification ISODATA a été réalisée, produisant 25 classes, et mettant en évidence les différents milieux forestiers des tropiques humides à une résolution spatiale de 1 km. Pour valider la classification, une méthode est proposée utilisant des images hautes résolutions Landsat 7 ETM+. Cette méthode permet de cadrer les vérifications sur le terrain et d'approfondir la définition des classes de forêt tropicale humide. L'échantillon utilisé pour tester la méthode met en évidence les limites et contraintes de la démarche proposée. Ces points sont exprimés dans la discussion. Enfin les perspectives de développements sont abordées dans la conclusion.
https://hal.archives-ouvertes.fr/hal-01189195

hal-01189195
https://hal.archives-ouvertes.fr/hal-01189195
PRODINRA : 13538

Éditeur(s) : HAL CCSD
Résumé : This paper presents the experimental results obtained for an automatic detection of ships in High Resolution optical satellite images. Our images are panchromatic SPOT 5 images, whose resolution is 5m per pixel. Our detection method is part of an industrial project of maritime surveillance. It is based on the Bayesian decision theory and does not need any preprocessing.The detection is composed of three stages. The first one is a pre-detection of targets that gives us candidates. The second one is a precise segmentation of each candidate. The third one is a classification of candidates into three classes : real small targets, real big targets and false alarms. The two first stages are based on the bayesian theory, using a very simple image model that leads to very fast algorithms. For now, the classification is based on multivariate decision tree. Finally, the overall results of the method are given for a set of images, as close as possible to the operational conditions and show the performances of the proposed method (detection of almost all the ships, false alarms due to crests of waves and an acceptable computing time).
Journée Nationale de Photogrammétrie et Télédétection

hal-00601854
https://hal.archives-ouvertes.fr/hal-00601854

Résumé : Les méthodes classiques d'analyse d'images satellites sont inadaptées au volume actuel du flux de données. L'automatisation de l'interprétation de ces images devient donc cruciale pour l'analyse et la gestion des phénomènes observables par satellite et évoluant dans le temps et l'espace. Ce travail vise à automatiser la cartographie dynamique de l'occupation du sol à partir d'images satellites, par des mécanismes expressifs, facilement interprétables en prenant en compte les aspects structurels de l'information géographique. Il s'inscrit dans le cadre de l'analyse d'images basée objet. Ainsi, un paramétrage supervisé d'un algorithme de segmentation d'images est proposé. Dans un deuxième temps, une méthode de classification supervisée d'objets géographiques est présentée combinant apprentissage automatique par programmation logique inductive et classement par l'approche multi-class rule set intersection. Ces approches sont appliquées à la cartographie de la bande côtière Guyanaise. Les résultats démontrent la faisabilité du paramétrage de la segmentation, mais également sa variabilité en fonction des classes de la carte de référence et des données d'entrée. Les résultats de la classification supervisée montrent qu'il est possible d'induire des règles de classification expressives, véhiculant des informations cohérentes et structurelles dans un contexte applicatif donnée et conduisant à des valeurs satisfaisantes de précision et de KAPPA (respectivement 84,6% et 0,7). Ce travail de thèse contribue ainsi à l'automatisation de la cartographie dynamique à partir d'images de télédétection et propose des perspectives originales et prometteuses.
Classical methods for satellite image analysis are inadequate for the current bulky data flow. Thus, automate the interpretation of such images becomes crucial for the analysis and management of phenomena changing in time and space, observable by satellite. Thus, this work aims at automating land cover cartography from satellite images, by expressive and easily interpretable mechanism, and by explicitly taking into account structural aspects of geographic information. It is part of the object-based image analysis framework, and assumes that it is possible to extract useful contextual knowledge from maps. Thus, a supervised parameterization methods of a segmentation algorithm is proposed. Secondly, a supervised classification of geographical objects is presented. It combines machine learning by inductive logic programming and the multi-class rule set intersection approach. These approaches are applied to the French Guiana coastline cartography. The results demonstrate the feasibility of the segmentation parameterization, but also its variability as a function of the reference map classes and of the input data. Yet, methodological developments allow to consider an operational implementation of such an approach. The results of the object supervised classification show that it is possible to induce expressive classification rules that convey consistent and structural information in a given application context and lead to reliable predictions, with overall accuracy and Kappa values equal to, respectively, 84,6% and 0,7. In conclusion, this work contributes to the automation of the dynamic cartography from remotely sensed images and proposes original and promising perpectives

http://www.theses.fr/2013AGUY0671/document

Éditeur(s) : HAL CCSD Society for the Study of Reproduction
Résumé : International audience
Matrix-assisted laser desorption/ionization (MALDI) tissue imaging mass spectrometry is particularly promising among the numerous applications of mass spectrometry. It is used for probing and analyzing the spatial arrangement of a wide range of molecules, including proteins, peptides, lipids, drugs, and metabolites, directly in thin slices of tissue. In the field of proteomics, the technology avoids tedious and time-consuming extraction and fractionation steps classically required for sample analysis. MALDI imaging mass spectrometry is increasingly recognized as a powerful method for clinical proteomics, particularly in cancer research. The technology has particular potential for the discovery of new tissue biomarker candidates, classification of tumors, early diagnosis or prognosis, elucidating pathogenesis pathways, and therapy monitoring. Over recent years, MALDI imaging mass spectrometry has been used for molecular profiling and imaging directly in male and female reproductive tissues. This review will consider some of the recent publications in the field, addressing a range of issues covering embryo development, gene expression product profiling during gametogenesis, and seeking and identifying biomarkers of reproductive cancers. The wealth of advances in mass spectrometry imaging will inevitably attract biologists and clinicians as the advantages and power of this technology become more widely known. This review will also discuss bottlenecks and the many technical issues that remain to be resolved before laboratories in the field can adopt the technology. We foresee that MALDI imaging mass spectrometry will have a major impact in reproductive research by opening new avenues to the understanding of various molecular mechanisms and the diagnosis of reproductive pathologies.
ISSN: 0006-3363

hal-00877851
https://hal.archives-ouvertes.fr/hal-00877851
DOI : 10.1095/biolreprod.111.094896
PUBMED : 22156474

Éditeur(s) : HAL CCSD
Résumé : Weather stations are constantly provided with thousands of satellite images with a view to anticipating the progress of natural phenomenons. This constant flow ofinformation requires speed and a strict analysis. Consequently, it is absolutely necessary to implement systems capable of working night and day, in order to make assessments. In this article, we suggest a method relying on pattern recognition investigation techniques, such as segmentation and classification. These techniques will be used to locate and discriminate the banks of clouds coming from images picked up by the GOES geostationary satellite. Treatment of the cloud shapes which have been found can be carried out, using operations, such as mathematical morphology, so as to draw discriminating characteristics. Classifier systems are used in order to interpret the characteristics which have been obtained and to conduct cloud classification.
15th Meeting of the Caribbean Academy of Sciences
Gosier, Guadeloupe

hal-00524903
https://hal.archives-ouvertes.fr/hal-00524903

Éditeur(s) : HAL CCSD
Résumé : The research presented in this thesis aims at determining the possibility of partial or total automatic ship detection in High Resolution optical satellite images. At first, we expose the industrial context of our study, which fixes the constraints in terms of performances and processing time. Then, we present the available satellite data for our study (SPOT 5 panchromatic images of 5m per pixel resolution). Then we detail our approach for ship detection : the first stage is a pre-detection of targets that gives us candidates. The second stage is a precise segmentation of each candidate and the third stage is a classification of candidates. Our pre-detection method is based on the Bayesian decision theory. This pre-detection supplies a set of regions of interests (ROI). Among them, we have ships, but also false alarms due to clouds and to crests of waves. The purpose of the following stage is to eliminate the false alarms and to characterize the pre-detected ships. Then, we detail the method used to segment candidates. It uses active contours and also uses the Bayesian decision theory. After a brief description of the characteristics extracted for the classication of ships, we present the classifier that we used. We complete the ship characterization by presenting two methods of wave ship detection. The first approach uses active contours and the second is based on an a contrario approach. We compare the results of these two approaches. To complete the study and replace it in the industrial and operational context, we present experiments of the complete algorithm on complete images to estimate the performances and the processing time of the whole method.
Les travaux présentés dans cette thèse ont pour but de déterminer la possibilité d'automatiser, partiellement ou totalement, la détection de bateaux à partir d'images satellitaires optiques haute résolution. Nous exposons tout d'abord le contexte industriel dans lequelle s'insère notre étude, qui fixe les contraintes en termes de performances et de temps de traitement. Puis, nous présentons les données satellitaires disponibles dans le cadre de notre étude (images SPOT 5 panchromatique de résolution 5m). Après avoir illustré le contenu des scènes, nous présentons la démarche adoptée pour la détection des navires : une phase de pré-détection suivie d'une étape de segmentation et d'une étape de classification. La méthode de pré-détection utilisée s'appuie sur la théorie bayésienne de la décision. Après avoir présenté le modèle classique, nous discutons du modèle que nous avons utilisé. Cette pré-détection fournit un ensemble de régions d'intérêts. Parmi elles, nous n'avons pas uniquement des bateaux, nous avons également des fausses alarmes dues aux nuages et aux crêtes de vagues. Le but de l'étape suivante est à la fois d'éliminer les fausses alarmes et de caractériser les bateaux pré-détectés. Ensuite, nous détaillons la méthode utilisée pour segmenter les bateaux. Elle s'appuie sur des contours actifs et utilise également la théorie bayésienne de la décision. Après avoir décrit les caractéristiques extraites pour la classification des bateaux, nous présentons le classifieur utilisé. Nous complétons la caractérisation des navires en présentant deux méthodes de détection des sillages différentes des méthodes classiques. La première approche s'appuie sur l'utilisation de contours actifs et la seconde est basée sur une approche a contrario. Nous comparons les résultats des ces deux approches. An de compléter l'étude et se replacer dans le contexte industriel et opérationnel de la thèse, nous présentons des expérimentations de l'ensemble de l'algorithme sur des images complètes pour évaluer les performances et le temps de traitement.
https://tel.archives-ouvertes.fr/tel-00560119

tel-00560119
https://tel.archives-ouvertes.fr/tel-00560119
https://tel.archives-ouvertes.fr/tel-00560119/document
https://tel.archives-ouvertes.fr/tel-00560119/file/These_NadiaProia_Version_Finale.pdf

Éditeur(s) : HAL CCSD
Résumé : International audience
We propose in this article a color image segmentation technique based on an optimization of the Mean Shift method. It consists in classifying clusters of data points of a digital image, it does not require any preliminary designation of the number of classes and of their centers. The Mean Shift method applied to a digital color image creates a new image made up of aggregates of points belonging to a finished number of color classes. The complexity of this method based on a global approach of the image is in O(NxN), N being the number of pixels of the image. Our idea consists in applying a change of scale to the image to be segmented, to reduce the quantity of information and, in using a median filter to decrease the number of colors to minimize the complexity of the latter. The comparative study which we present shows that the optimization which we proposed gives better, reliable results than the classic use of the Mean Shift method.
Complex, Intelligent and Software Intensive Systems, International Conference
Cracovie, Poland

hal-00508427
https://hal.archives-ouvertes.fr/hal-00508427
DOI : 10.1109/CISIS.2010.62

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
The number of satellites and sensors devoted to Earth observation has become increasingly elevated, delivering extensive data, especially images. At the same time, the access to such data and the tools needed to process them has considerably improved. In the presence of such data flow, we need automatic image interpretation methods, especially when it comes to the monitoring and prediction of environmental and societal changes in highly dynamic socio-environmental contexts. This could be accomplished via artificial intelligence. The concept described here relies on the induction of classification rules that explicitly take into account structural knowledge, using Aleph, an Inductive Logic Programming (ILP) system, combined with a multi-class classification procedure. This methodology was used to monitor changes in land cover/use of the French Guiana coastline. One hundred and fifty-eight classification rules were induced from 3 diachronic land cover/use maps including 38 classes. These rules were expressed in first order logic language, which makes them easily understandable by non-experts. A 10-fold cross-validation gave significant average values of 84.62%, 99.57% and 77.22% for classification accuracy, specificity and sensitivity, respectively. Our methodology could be beneficial to automatically classify new objects and to facilitate object-based classification procedures.
ISSN: 0098-3004

hal-01369822
https://hal.archives-ouvertes.fr/hal-01369822
https://hal.archives-ouvertes.fr/hal-01369822/document
https://hal.archives-ouvertes.fr/hal-01369822/file/CAGEO-D-13-00538R1_article.pdf
DOI : 10.1016/j.cageo.2014.08.013

Éditeur(s) : HAL CCSD American Chemical Society
Résumé : International audience
Childhood absence epilepsy is a prototypic form of generalized nonconvulsive epilepsy characterized by short impairments of consciousness concomitant with synchronous and bilateral spike-and-wave discharges in the electroencephalogram. For scientists in this field, the BS/Orl and BR/Orl mouse lines, derived from a genetic selection, constitute an original mouse model "in mirror" of absence epilepsy. The potential of MALDI imaging mass spectrometry (IMS) for the discovery of potential biomarkers is increasingly recognized. Interestingly, statistical analysis tools specifically adapted to IMS data sets and methods for the identification of detected proteins play an essential role. In this study, a new cross-classification comparative design using a combined discrete wavelet transformation-support vector machine classification was developed to discriminate spectra of brain sections of BS/Orl and BR/Orl mice. Nineteen m/z ratios were thus highlighted as potential markers with very high recognition rates (87-99%). Seven of these potential markers were identified using a top-down approach, in particular a fragment of Synapsin-I. This protein is yet suspected to be involved in epilepsy. Immunohistochemistry and Western Blot experiments confirmed the differential expression of Synapsin-I observed by IMS, thus tending to validate our approach. Functional assays are being performed to confirm the involvement of Synapsin-I in the mechanisms underlying childhood absence epilepsy.
ISSN: 1535-3893

hal-00877728
https://hal.archives-ouvertes.fr/hal-00877728
DOI : 10.1021/pr3006974
PUBMED : 22994238