Automatic Surveillance Vision Detection Using Gaussian...

Automatic surveillance application has been a subject of extensive research during last decades to develop robust application. As one of its important application, the vision detection and localization interest in detecting moving objects under the challenging conditions of the illumination change, occlusion, shadow and perturbation of the images sources. In vision detection system, the techniques used are either background subtraction based or feature detection based. The background subtraction based techniques interests in detecting variation within the scene across several image frames. This approach is based on comparing the current image with a reference one(s) of the background. Pixels of sharp variations are consequently classified†¦show more content†¦Another technique is the Eigen backgrounds [6], this approach is based on an eigenvalue decomposition which is applied to the whole image. This extended spatial domain enabled the exploration of the spatial correlation and avoiding the tiling effect of block partitioning. The method has two phases that alternate with a learning phase for samples acquisition and eigenvector matrix computation and followed by classification step for foreground detection. The Gaussian process presented in [7] , is a non parametric tool that performs Bayesian inference, is gaining great respect of the research community, and thought most of the investigation in this domain is focused on its theory, great efforts consider interpreting this theory to real application. In the same reference, the binary Gaussian process classifier is proposed as foundation of Gaussian process classiï ¬ cation. While in [8], Laplace approximation as a tool of inference in Gaussian process models was suggested as a tool supporting the Gaussian binary classifier. The generalization from two-class classification using binary Gaussian classifier to multiclass classification was proposed in [9], in this approach, a technique of voting and combinations of approximate posterior probabilities is used to achieve simpleShow MoreRelatedGait Analysis8133 Words   |  33 PagesAbstract ABSTRACT One of the main goals of computer vision research is to develop methods for recognition of objects and events. A subclass of these problems is the recognition of humans and their activities. Recognition of humans from arbitrary viewpoints is an important requirement for different applications such as intelligent environments, surveillance and access control. Human gait is an attractive modality for recognizing people at a distance. Human gait is a spatio-temporal phenomenon