Abstract
In this thesis, we propose a statistical model-based contour tracking method based on the Condensation framework. The models include a novel contour prediction model and two statistical object models. The object models consist of the grayscale histogram and contour shape PCA models computed from the previous tracking results. With the incremental singular value decomposition (SVD) technique, these three models are learned and updated very efficiently during tracking. We show that the proposed shape prediction model performs better than the affine predictor though experiments. Experimental results show the proposed contour tracking algorithm is very stable in tracking human heads on real videos with object scaling, rotation, partial occlusion, and illumination changes.