Stochastic visual tracking with active appearance models

[摘要] ENGLISH ABSTRACT: In many applications, an accurate, robust and fast tracker is needed, for example in surveillance,gesture recognition, tracking lips for lip-reading and creating an augmented reality by embeddinga tracked object in a virtual environment. In this dissertation we investigate the viability of atracker that combines the accuracy of active appearancemodels with the robustness of the particle lter (a stochastic process)-we call this combination the PFAAM. In order to obtain a fast system,we suggest local optimisation as well as using active appearance modelstted with non-linearapproaches.Active appearance models use both contour (shape) and greyscale information to build adeformable template of an object. ey are typically accurate, but not necessarily robust, whentracking contours. A particlelter is a generalisation of the Kalmanlter. In a tutorial style,we show how the particlelter is derived as a numerical approximation for the general stateestimation problem.e algorithms are tested for accuracy, robustness and speed on a PC, in an embeddedenvironment and by tracking in ìD. e algorithms run real-time on a PC and near real-time inour embedded environment. In both cases, good accuracy and robustness is achieved, even if thetracked object moves fast against a cluttered background, and for uncomplicated occlusions.

[发布机构] Stellenbosch University