Urban Detection From Hyperspectral Images Using Dimension-Reduction Model and Fusion of Multiple Segmentations Based on Stuctural and Textural Features

Ce mémoire de maîtrise présente une nouvelle approche non supervisée pour détecter et segmenter les régions urbaines dans les images hyperspectrales. La méthode proposée n ́ecessite trois étapes. Tout d’abord, afin de réduire le coût calculatoire de notre algorithme, une image couleur du contenu spectral est estimée. A cette fin, une étape de réduction de dimensionalité non-linéaire, basée sur deux critères complémentaires mais contradictoires de bonne visualisation; à savoir la précision et le contraste, est réalisée pour l’affichage couleur de chaque image hyperspectrale. Ensuite, pour discriminer les régions urbaines des régions non urbaines, la seconde étape consiste à extraire quelques caractéristiques discriminantes (et complémentaires) sur cette image hyperspectrale couleur. A cette fin, nous avons extrait une série de paramètres discriminants pour décrire les caractéristiques d’une zone urbaine, principalement composée d’objets manufacturés de formes simples g ́eométriques et régulières. Nous avons utilisé des caractéristiques texturales basées sur les niveaux de gris, la magnitude du gradient ou des paramètres issus de la matrice de co-occurrence combinés avec des caractéristiques structurelles basées sur l’orientation locale du gradient de l’image et la détection locale de segments de droites. Afin de réduire encore la complexité de calcul de notre approche et éviter le problème de la ”malédiction de la dimensionnalité” quand on décide de regrouper des données de dimensions élevées, nous avons décidé de classifier individuellement, dans la dernière étape, chaque caractéristique texturale ou structurelle avec une simple procédure de K-moyennes et ensuite de combiner ces segmentations grossières, obtenues à faible coût, avec un modèle efficace de fusion de cartes de segmentations. Les expérimentations données dans ce rapport montrent que cette stratégie est efficace visuellement et se compare favorablement aux autres méthodes de détection et segmentation de zones urbaines à partir d’images hyperspectrales.

Photorealistic Surface Rendering with Microfacet Theory

La synthèse d'images dites photoréalistes nécessite d'évaluer numériquement la manière dont la lumière et la matière interagissent physiquement, ce qui, malgré la puissance de calcul impressionnante dont nous bénéficions aujourd'hui et qui ne cesse d'augmenter, est encore bien loin de devenir une tâche triviale pour nos ordinateurs. Ceci est dû en majeure partie à la manière dont nous représentons les objets: afin de reproduire les interactions subtiles qui mènent à la perception du détail, il est nécessaire de modéliser des quantités phénoménales de géométries. Au moment du rendu, cette complexité conduit inexorablement à de lourdes requêtes d'entrées-sorties, qui, couplées à des évaluations d'opérateurs de filtrage complexes, rendent les temps de calcul nécessaires à produire des images sans défaut totalement déraisonnables. Afin de pallier ces limitations sous les contraintes actuelles, il est nécessaire de dériver une représentation multiéchelle de la matière. Dans cette thèse, nous construisons une telle représentation pour la matière dont l'interface correspond à une surface perturbée, une configuration qui se construit généralement via des cartes d'élévations en infographie. Nous dérivons notre représentation dans le contexte de la théorie des microfacettes (conçue à l'origine pour modéliser la réflectance de surfaces rugueuses), que nous présentons d'abord, puis augmentons en deux temps. Dans un premier temps, nous rendons la théorie applicable à travers plusieurs échelles d'observation en la généralisant aux statistiques de microfacettes décentrées. Dans l'autre, nous dérivons une procédure d'inversion capable de reconstruire les statistiques de microfacettes à partir de réponses de réflexion d'un matériau arbitraire dans les configurations de rétroréflexion. Nous montrons comment cette théorie augmentée peut être exploitée afin de dériver un opérateur général et efficace de rééchantillonnage approximatif de cartes d'élévations qui (a) préserve l'anisotropie du transport de la lumière pour n'importe quelle résolution, (b) peut être appliqué en amont du rendu et stocké dans des MIP maps afin de diminuer drastiquement le nombre de requêtes d'entrées-sorties, et (c) simplifie de manière considérable les opérations de filtrage par pixel, le tout conduisant à des temps de rendu plus courts. Afin de valider et démontrer l'efficacité de notre opérateur, nous synthétisons des images photoréalistes anticrenelées et les comparons à des images de référence. De plus, nous fournissons une implantation C++ complète tout au long de la dissertation afin de faciliter la reproduction des résultats obtenus. Nous concluons avec une discussion portant sur les limitations de notre approche, ainsi que sur les verrous restant à lever afin de dériver une représentation multiéchelle de la matière encore plus générale.

Réduction des artéfacts de tuteur coronarien au moyen d’un algorithme de reconstruction avec renforcement des bords : étude prospective transversale en tomodensitométrie 256 coupes

Les artéfacts métalliques entraînent un épaississement artéfactuel de la paroi des tuteurs en tomodensitométrie (TDM) avec réduction apparente de leur lumière. Cette étude transversale prospective, devis mesures répétées et observateurs avec méthode en aveugle, chez 24 patients consécutifs/71 tuteurs coronariens a pour objectif de comparer l’épaisseur de paroi des tuteurs en TDM après reconstruction par un algorithme avec renforcement des bords et un algorithme standard. Une angiographie coronarienne par TDM 256 coupes a été réalisée, avec reconstruction par algorithmes avec renforcement des bords et standard. L’épaisseur de paroi des tuteurs était mesurée par méthodes orthogonale (diamètres) et circonférentielle (circonférences). La qualité d’image des tuteurs était évaluée par échelle ordinale, et les données analysées par modèles linéaire mixte et régression logistique des cotes proportionnelles. L’épaisseur de paroi des tuteurs était inférieure avec l’algorithme avec renforcement des bords comparé à l’algorithme standard, avec les méthodes orthogonale (0,97±0,02 vs 1,09±0,03 mm, respectivement; p<0,001) et circonférentielle (1,13±0,02 vs 1,21±0,02 mm, respectivement; p<0,001). Le premier causait moins de surestimation par rapport à l’épaisseur nominale comparé au second, avec méthodes orthogonale (0,89±0,19 vs 1,00±0,26 mm, respectivement; p<0,001) et circonférentielle (1,06±0,26 vs 1,13±0,31 mm, respectivement; p=0,005) et diminuait de 6 % la surestimation. Les scores de qualité étaient meilleurs avec l’algorithme avec renforcement des bords (OR 3,71; IC 95% 2,33–5,92; p<0,001). En conclusion, la reconstruction des images avec l’algorithme avec renforcement des bords génère des parois de tuteurs plus minces, moins de surestimation, et de meilleurs scores de qualité d’image que l’algorithme standard.

Theory of Breaking of a Potyrner Molecule under Tension

The thesis presents the dynamics of a polymer chain under tension. It includes existing theories of polymer fracture, important theories of reaction rates, the rate using multidimensional transition state theory and apply it to the case of polyethylene etc. The main findings of the study are; the life time of the bond is somewhat sensitive to the potential lead to rather different answers, for a given potential a rough estimate of the rate can be obtained by a simples approximation that considers the dynamics of only the bond that breaks and neglects the coupling to neighboring bonds. Dynamics of neighboring bonds would decrease the rate, but usually not more than by one order of magnitude, for the breaking of polyethylene, quantum effects are important only for temperatures below 150K, the lifetime strongly depends on the strain and as the strain varies over a narrow range, the life varies rapidly from 105 seconds to 10_5 seconds, if we change one unit of the polymer by a foreign atom, say by one sulphure atom, in the main chain itself, by a weaker bond, the rate is found to increase by orders of magnitude etc.

Statistical Machine Learning Techniques for the Prediction of Learning Disabilities in School-Age Children

Learning Disability (LD) is a general term that describes specific kinds of learning problems. It is a neurological condition that affects a child's brain and impairs his ability to carry out one or many specific tasks. The learning disabled children are neither slow nor mentally retarded. This disorder can make it problematic for a child to learn as quickly or in the same way as some child who isn't affected by a learning disability. An affected child can have normal or above average intelligence. They may have difficulty paying attention, with reading or letter recognition, or with mathematics. It does not mean that children who have learning disabilities are less intelligent. In fact, many children who have learning disabilities are more intelligent than an average child. Learning disabilities vary from child to child. One child with LD may not have the same kind of learning problems as another child with LD. There is no cure for learning disabilities and they are life-long. However, children with LD can be high achievers and can be taught ways to get around the learning disability. In this research work, data mining using machine learning techniques are used to analyze the symptoms of LD, establish interrelationships between them and evaluate the relative importance of these symptoms. To increase the diagnostic accuracy of learning disability prediction, a knowledge based tool based on statistical machine learning or data mining techniques, with high accuracy,according to the knowledge obtained from the clinical information, is proposed. The basic idea of the developed knowledge based tool is to increase the accuracy of the learning disability assessment and reduce the time used for the same. Different statistical machine learning techniques in data mining are used in the study. Identifying the important parameters of LD prediction using the data mining techniques, identifying the hidden relationship between the symptoms of LD and estimating the relative significance of each symptoms of LD are also the parts of the objectives of this research work. The developed tool has many advantages compared to the traditional methods of using check lists in determination of learning disabilities. For improving the performance of various classifiers, we developed some preprocessing methods for the LD prediction system. A new system based on fuzzy and rough set models are also developed for LD prediction. Here also the importance of pre-processing is studied. A Graphical User Interface (GUI) is designed for developing an integrated knowledge based tool for prediction of LD as well as its degree. The designed tool stores the details of the children in the student database and retrieves their LD report as and when required. The present study undoubtedly proves the effectiveness of the tool developed based on various machine learning techniques. It also identifies the important parameters of LD and accurately predicts the learning disability in school age children. This thesis makes several major contributions in technical, general and social areas. The results are found very beneficial to the parents, teachers and the institutions. They are able to diagnose the child’s problem at an early stage and can go for the proper treatments/counseling at the correct time so as to avoid the academic and social losses.

Packaging materials for shrimp, fish and fish products, their properties, selection and effect of different packaging materials on their shelf life

Packaging is important not only in extending the shellife of fish and fishery products but also improving their marketability. In the recent years, significant development have taken place in the packaging industry. During the past decade in India, there is almost a packaging revolution with the availability of variety packaging materials, thus generating better packaging consciousness in other producer/manufacturing industries. But unfortunately, such realisation is not forthcoming in the fisheries sector and packaging techniques for local and export trade continues to be on traditional lines with their inherent drawbacks and limitations. Better packaging ensures improved quality and presentation of the products and ensures higher returns to the producer. Among several packaging materials used in fishery industry, ISI specifications had been formulated only for corrugated fibre board boxes for export of seafoods and froglegs. This standard was formulated before containersiation came into existance in the export of marine products. Before containerisation, the standards were stringent in view of the rough handling, transportation and storage. Two of the common defects reported in the master cartons exported from India are low mechanical strength and tendency to get wet. They are weakened by the deposits of moisture caused by temperature fluctuations during loading, unloading and other handling stages. It is necessary to rectify the above defects in packaging aquatic products and hence in the present study extensive investigations are carried out to find out the reasons for the damage of master cartons, to evolve code of practice for the packaging oi frozen shrimp for exports, development of alternative style of packaging for the shipping container, development of suitable consumer packaging materials for fish soup powder, cured dried mackeral, fish pickles in oil and frozen shrimp. For the development of suitable packaging materials, it is absolutely essential to know the properties of packaging materials, effect of different packaging materials on theirshelf life and their suitability for food contact applications.

A study of the cost structure of the cashew industry in kerala, karnataka and tamilnadu

Since the world demand for cashew kernels has been rising steadily for several years in the past, conferring significant price increase the processing of cashew remains a highly profitable lndustry. India being the earliest and largest supplier of cashew kernels in the world market it is our prestigious obligations to reestablish her pristine monopoly. Further the added importance ot the indutry in the Socio economic context of the State of Kerala makes various measures impervative in order to bring back to the industry its pristine glory at the late sixties to give a face lift and to stabilize the industry. This present study adopts a comprehensive frame work of analysis compassing the major issues involved in the cultivation, distribution, import, processing and marketing of cashew undcr the private and public sector, migration of the industry and the financial requirements of the industry.

Grain boundaries in vortex matter

We explore the statistical properties of grain boundaries in the vortex polycrystalline phase of type-II superconductors. Treating grain boundaries as arrays of dislocations interacting through linear elasticity, we show that self-interaction of a deformed grain boundary is equivalent to a nonlocal long-range surface tension. This affects the pinning properties of grain boundaries, which are found to be less rough than isolated dislocations. The presence of grain boundaries has an important effect on the transport properties of type-II superconductors as we show by numerical simulations: our results indicate that the critical current is higher for a vortex polycrystal than for a regular vortex lattice. Finally, we discuss the possible role of grain boundaries in vortex lattice melting. Through a phenomenological theory we show that melting can be preceded by an intermediate polycrystalline phase.

Studies On Sediment Transport In The Surf Zone Along Certain Beaches Of Kerala

Sediment transport in the nearshore areas is an important process in deciding the coastline stability. The design and effective maintenance of navigable waterways, harbours and marine structures depend on the stability of the sediment substrate and the nature of sedimentation in the nearshore zone. The nearshore zone is a complex environment and the exact relationships existing between water motions and the resulting sediment transports are not well understood. During the rough weather season, when the sediment movement is considerable, processes occurring in the nearshore area are much less understood. Moreover, there is a general lack of field measurements, especially during the time of severe storm conditions. The increasing pressures and the concern on the preservation of the valuable coastal environment have led to the development of shore protection programmes. Conservation not only demands knowledge of what needs to be done, but also requires the basic processes to be fully understood. Considering the fragile nature of barrier beaches and intense occupancy of these areas by man, these coastal features have long been a subject of study by coastal oceanographers, geomorphologists and engineers. The present study is an attempt to understand the sediment movement in relation to beach dynamics, especially in the surf zone, along some part of Kerala coast and the response of the beaches to various forcing functions over different seasons

Design of an Edge -coupled dual -ring Split-ring resonator

Electric permittivity and magnetic permeability control electromagnetic wave propagation th rough materials. I n naturally occu rring materials, these are positive. Artificial materials exhi b iting negative material properties have been reported : they are referred to as metamaterials. This paper concentrates on a ring-type split-ring resonator (SRR) exhibiting negative magnetic permeability. The design and synthesis of the SRR using the genetic-algorithm approach is explained in detail. A user-friendly g raphical user i nterface (G U I ) for an SRR optim izer and estimator using MATLAB TM is also presented

Influence of substrate topography on the growth and magnetic properties of obliquely deposited amorphous nanocolumns of Fe–Ni

We investigated the influence of substrate surface roughness on the structural and magnetic properties of obliquely deposited amorphous nanocolumns of Fe–Ni. Experiments showed that the surface roughness of the substrate greatly determines the morphology of the columnar structures and this in turn has a profound influence on the magnetic properties. Nucleation of Fe–Ni nanocolumns on a smooth silicon substrate was at random, while that on a rough glass substrate was defined by the irregularities on the substrate surface. It has been found that magnetic interaction between the nanocolumns prepared on a silicon substrate was due to their small inter-column separation. Well separated nanocolumns on a glass substrate resulted in exchange isolated magnetic domains. The size, shape and the distribution of nanocolumns can be tailored by appropriately choosing the surface roughness of the substrate. This will find potential applications in thin film magnetism.

Error estimates for approximate approximations on compact intervals

The aim of this paper is the investigation of the error which results from the method of approximate approximations applied to functions defined on compact in- tervals, only. This method, which is based on an approximate partition of unity, was introduced by V. Mazya in 1991 and has mainly been used for functions defied on the whole space up to now. For the treatment of differential equations and boundary integral equations, however, an efficient approximation procedure on compact intervals is needed. In the present paper we apply the method of approximate approximations to functions which are defined on compact intervals. In contrast to the whole space case here a truncation error has to be controlled in addition. For the resulting total error pointwise estimates and L1-estimates are given, where all the constants are determined explicitly.

Untersuchungen zum Gesundheitsstatus und zu betrieblichen Maßnahmen der Gesundheitsvorsorge in der ökologischen Schweinehaltung

Im Rahmen dieser Arbeit wurde mittels einer umfassenden und von einer Projektgruppe begleiteten Literaturanalyse sowie anhand von Praxiserhebungen der Frage nachgegangen, wie es um die Tiergesundheit in der ökologischen Schweinehaltung bestellt ist. Ferner wurde untersucht, ob die ökologischen Rahmenrichtlinien (EWG-VO 2092/91) ein Hemmnis für die Sicherstellung eines hohen Gesundheitsstatus der Nutztiere darstellen. Die Literaturauswertung ergab, dass die Erkrankungsraten sowohl auf ökologisch wie auf konventionell geführten Betrieben ein hohes Niveau aufwiesen. Die Ursachen wurden vorrangig in einem unzureichenden betrieblichen Management gesehen, während den EG-Rahmenrichtlinien in der Regel keine maßgebliche Bedeutung für den Gesundheitsstatus beigemessen wurde. In Praxiserhebungen wurde die Status-quo-Situation zur Tiergesundheit auf sechs ökologisch wirt- schaftenden Schweinemastbetrieben erfasst. Ferner wurde untersucht, inwieweit eine Verbesserung des Gesundheitsstatus auf den Betrieben mit einem Mehraufwand an Arbeitszeit und Investitionen verbunden ist. Hierzu wurden die Bereiche Haltung, Fütterung und Hygienemanagement mittels einer Schwachstellenanalyse (Hazard Analysis Critical Control Points-Konzept) beurteilt. Der Zeit- bedarf für Maßnahmen zur Gesundheitsvorsorge wurde durch Arbeitszeitaufzeichnungen erfasst. Die Untersuchungen zeigten, dass auf den Betrieben zum Teil erhebliche Mängel im Hygiene- management bestanden. Anhand von Schlachtkörperbefunden wurde eine hohe Rate pathologischer Leberveränderungen, verursacht durch Larven des Schweinespulwurms Ascaris suum, diagnosti- ziert. Insgesamt wiesen die Schlachtkörper der Tiere von den sechs Betrieben mit durchschnittlich 48,8 % sowie die Schlachtkörper von ökologischen Vergleichsbetrieben mit 63,9 % signifikant höhere Befundraten auf als die Schlachtkörper konventioneller Vergleichsbetriebe (17,4 %). Der Arbeitszeitbedarf für die Mastschweinehaltung auf den sechs Betrieben variierte zwischen 2,0 und 3,7 Arbeitskraftstunden pro Mastplatz und Jahr und lag damit deutlich höher als vergleichbare Literaturdaten. Zwischen den Betrieben traten hohe Unterschiede hinsichtlich der zeitlichen Aufwendungen für Hygiene- und Tiergesundheitsmaßnahmen auf. Während der Interventionsphase wurden das Entwurmungs- und das Hygieneregime der sechs Betriebe optimiert. Die Buchten wurden ordnungsgemäß gereinigt und desinfiziert und der erforderliche Arbeitszeitaufwand quantifiziert. Dieser lag zusammen mit den zusätzlichen Kosten um ca. 46 % höher als der betrieblich übliche Arbeitszeitaufwand. Betriebe mit neu konzipierten Ställen und glatten Buchtenböden und -wänden konnten diese mit vergleichsweise geringem Arbeitsaufwand reinigen und desinfizieren. In einigen Altgebäuden hingegen war aufgrund rauer und schadhafter Oberflächen eine ordnungsgemäße Reinigung und Desinfektion nur bedingt durchzuführen. Die Rate der Leberveränderungen am Schlachthof konnte von durchschnittlich 48,9 % auf 32,7 % gesenkt werden. Während sich die Befundraten auf einigen Betrieben signifikant verbesserten, trat bei der Nutzung von unrenovierten Ställen in Altgebäuden keine Besserung ein. Es wird geschlussfolgert, dass die Rahmenbedingungen der EG-Verordnung einem hohen Tiergesundheitsstatus nicht entgegen stehen, in der Regel jedoch mit einem erhöhten Aufwand für Hygienemaßnahmen verbunden sind. Voraussetzung für eine geringe Parasitenbelastung sind ein konsequent umgesetztes Hygiene- und Entwurmungsregime sowie gut zu reinigende Buchten. Letzteres ist allerdings bei der Nutzung von Altgebäuden nicht immer gegeben.

Notes on PCA, Regularization, Sparsity and Support Vector Machines

We derive a new representation for a function as a linear combination of local correlation kernels at optimal sparse locations and discuss its relation to PCA, regularization, sparsity principles and Support Vector Machines. We first review previous results for the approximation of a function from discrete data (Girosi, 1998) in the context of Vapnik"s feature space and dual representation (Vapnik, 1995). We apply them to show 1) that a standard regularization functional with a stabilizer defined in terms of the correlation function induces a regression function in the span of the feature space of classical Principal Components and 2) that there exist a dual representations of the regression function in terms of a regularization network with a kernel equal to a generalized correlation function. We then describe the main observation of the paper: the dual representation in terms of the correlation function can be sparsified using the Support Vector Machines (Vapnik, 1982) technique and this operation is equivalent to sparsify a large dictionary of basis functions adapted to the task, using a variation of Basis Pursuit De-Noising (Chen, Donoho and Saunders, 1995; see also related work by Donahue and Geiger, 1994; Olshausen and Field, 1995; Lewicki and Sejnowski, 1998). In addition to extending the close relations between regularization, Support Vector Machines and sparsity, our work also illuminates and formalizes the LFA concept of Penev and Atick (1996). We discuss the relation between our results, which are about regression, and the different problem of pattern classification.

A comparison of the alr and ilr transformations for kernel density estimation of compositional data

In a seminal paper, Aitchison and Lauder (1985) introduced classical kernel density estimation techniques in the context of compositional data analysis. Indeed, they gave two options for the choice of the kernel to be used in the kernel estimator. One of these kernels is based on the use the alr transformation on the simplex SD jointly with the normal distribution on RD-1. However, these authors themselves recognized that this method has some deficiencies. A method for overcoming these dificulties based on recent developments for compositional data analysis and multivariate kernel estimation theory, combining the ilr transformation with the use of the normal density with a full bandwidth matrix, was recently proposed in Martín-Fernández, Chacón and Mateu- Figueras (2006). Here we present an extensive simulation study that compares both methods in practice, thus exploring the finite-sample behaviour of both estimators