As digitisation era will come, a large number of multimedia datas (image, video, etc.) are stored in the database by digitisation, and its retrieval system is more and more important. Video is huge in frames amount, in order to search effectively and fast, the first step will detect and examine the place where the scene changes in the video, cut apart the scene, find out the key frame from each scene, regard as analysis that the index file searches with the key frame. The scene changes the way and divides into the abrupt and the gradual transition. But in the video, even if in the same scene, incident of often violent movements or the camera are moving etc. happens, and obscure with the gradual transition to some extent. Thus this papper gets the main component from every frame in the video using principal component analysis (PCA), reduce the noise to interfere, and classify these feauture points with support vector clustering, it is the same class that the close feature points is belonged to. If the feature points are located between two groups of different datas, represent the scene is changing slowly in the video, detect scene change by this.

Cette dissertation présente un ensemble d'algorithmes visant à en permettre un usage rapide, robuste et automatique des « Support Vector Machines » (SVM) non supervisés dans un contexte d'analyse de données. Les SVM non supervisés se déclinent sous deux types algorithmes prometteurs, le « Support Vector Clustering » (SVC) et le « Support Vector Domain Description » (SVDD), offrant respectivement une solution à deux problèmes importants en analyse de données, soit la recherche de groupements homogènes (« clustering »), ainsi que la reconnaissance d'éléments atypiques (« novelty/abnomaly detection ») à partir d'un ensemble de données. Cette recherche propose des solutions concrètes à trois limitations fondamentales inhérentes à ces deux algorithmes, notamment I) l'absence d'algorithme d'optimisation efficace permettant d'exécuter la phase d'entrainement des SVDD et SVC sur des ensembles de données volumineux dans un délai acceptable, 2) le manque d'efficacité et de robustesse des algorithmes existants de partitionnement des données pour SVC, ainsi que 3) l'absence de stratégies de sélection automatique des hyperparamètres pour SVDD et SVC contrôlant la complexité et la tolérance au bruit des modèles générés. La résolution individuelle des trois limitations mentionnées précédemment constitue les trois axes principaux de cette thèse doctorale, chacun faisant l'objet d'un article scientifique proposant des stratégies et algorithmes permettant un usage rapide, robuste et exempt de paramètres d'entrée des SVDD et SVC sur des ensembles de données arbitraires.

During the Digital Period, the more people using these digital video. When there are more and more users and amount of video data, the management of video data becomes a significant dimension during development. Therefore, there are more and more studying of accomplishing video database system, which provide users to search and get them. In this paper, a novel method for Video Scene Change Detection and video database retrieval is proposed. Uses Fractal orthonormal bases to guarantee the similar index has the similar image the characteristic union support vector clustering, splits a video into a sequence of shots, extracts a few representative frames(key-frames) to take the video database index from each shot. When image search compared to, according to MIL to pick up the characteristic, which images pursues the video database to have the similar characteristic, computation similar, makes the place output according to this.

AdaBoost.M1 has been successfully applied to improve the accuracy of a learning algorithm for multi-class classification problems. However, it assumes that the performance of each base classifier must be better than 1/2, and this may be hard to achieve in practice for a multi-class problem. A new algorithm called AdaBoost.MK only requiring base classifiers better than a random guessing (1/k) is thus designed. Early SVM-based multi-class classification algorithms work by splitting the original problem into a set of two-class sub-problems. The time and space required by these algorithms are very demanding. In order to have low time and space complexities, we develop a base classifier that integrates one-class SVMs with discriminant functions. In this study, a hybrid method that integrates AdaBoost.MK and one-class SVMs with improved discriminant functions as the base classifiers is proposed to solve a multi-class classification problem. Experimental results on data sets from UCI and Statlog show that the proposed approach outperforms many popular multi-class algorithms including support vector clustering and AdaBoost.M1 with one-class SVMs as the base classifiers.

Classification algorithms have been widely used in many application domains. Most of these domains deal with massive collection of data and hence demand classification algorithms that scale well with the size of the data sets involved. A classification algorithm is said to be scalable if there is no significant increase in time and space requirements for the algorithm (without compromising the generalization performance) when dealing with an increase in the training set size. Support Vector Machine (SVM) is one of the most celebrated kernel based classification methods used in Machine Learning. An SVM capable of handling large scale classification problems will definitely be an ideal candidate in many real world applications. The training process involved in SVM classifier is usually formulated as a Quadratic Programing(QP) problem. The existing solution strategies for this problem have an associated time and space complexity that is (at least) quadratic in the number of training points. This makes the SVM training very expensive even on classification problems having a few thousands of training examples. This thesis addresses the scalability of the training algorithms involved in both two class and multiclass Support Vector Machines. Efficient training schemes reducing the space and time requirements of the SVM training process are proposed as possible solutions. The classification schemes discussed in the thesis for handling large scale two class classification problems are a) Two selective sampling based training schemes for scaling Non-linear SVM and b) Clustering based approaches for handling unbalanced data sets with Core Vector Machine. To handle large scale multicalss classification problems, the thesis proposes Multiclass Core Vector Machine (MCVM), a scalable SVM based multiclass classifier. In MVCM, the multiclass SVM problem is shown to be equivalent to a Minimum Enclosing Ball (MEB) problem and is then solved using a fast approximate MEB finding algorithm. Experimental studies were done with several large real world data sets such as IJCNN1 and Acoustic data sets from LIBSVM page, Extended USPS data set from CVM page and network intrusion detection data sets of DARPA, US Defense used in KDD 99 contest. From the empirical results it is observed that the proposed classification schemes achieve good generalization performance at low time and space requirements. Further, the scalability experiments done with large training data sets have demonstrated that the proposed schemes scale well. A novel soft clustering scheme called Rough Support Vector Clustering (RSVC) employing the idea of Soft Minimum Enclosing Ball Problem (SMEB) is another contribution discussed in this thesis. Experiments done with a synthetic data set and the real world data set namely IRIS, have shown that RSVC finds meaningful soft cluster abstractions.

Made available in DSpace on 2016-03-15T19:37:30Z (GMT). No. of bitstreams: 1 Mauro Ulisses Schneider.pdf: 1455677 bytes, checksum: 6eba28391f8f6910fbf5457a57119bd3 (MD5) Previous issue date: 2010-08-23 / Fundo Mackenzie de Pesquisa / The use of biometric systems has been widely stimulated by both the government and private entities to replace or improve traditional security systems. Biometric systems are becoming increasingly indispensable to protecting life and property, mainly due to its robustness, reliability, difficult to counterfeit and fast authentication. In real world applications, the devices for image acquisition and the environment are not always controlled and may under certain circumstances produce noisy images or with large variations in tonality, texture, geometry, hindering segmentation and consequently the authentication of the an individual. To deal effectively with such problems, this dissertation investigates the possibility of using committee machines combined with digital image processing techniques for iris segmentation. The components employed in the composition of the committee machines are support vector clustering, k-means and self organizing maps. In order to evaluate the performance of the tools developed in this dissertation, the experimental results obtained are compared with related works reported in the literature. Experiments on publicity available UBIRIS database indicate that committee machine can be successfully applied to the iris segmentation. / A utilização de sistemas biométricos vem sendo amplamente; incentivados pelo governo e entidades privadas a fim de substituir ou melhorar os sistemas de segurança tradicionais. Os sistemas biométricos são cada vez mais indispensáveis para proteger vidas e bens, sendo robustos, confiáveis, de difícil falsificação e rápida autenticação. Em aplicações de mundo real, os dispositivos de aquisição de imagem e o ambiente nem sempre são controlados, podendo em certas circunstâncias produzir imagens ruidosas ou com grandes variações na tonalidade, textura, geometria, dificultando a sua segmentação e por conseqüência a autenticação do indivíduo. Para lidar eficazmente com tais problemas, nesta dissertação é estudado o emprego de comitês de máquinas em conjunto com técnicas de processamento de imagens digitais para a segmentação da íris. Os componentes estudados na composição do comitê de máquinas são agrupamento por vetores-suporte, k-means e mapas auto- organizáveis. Para a avaliação do desempenho das ferramentas desenvolvidas neste trabalho, os resultados obtidos são comparados com trabalhos relacionados na literatura. Foi utilizada a base de dados pública UBIRIS disponível na internet.