Human Body Tracking by Adaptive Background Models and Mean-Shift Analysis
Fatih Porikli, Oncel Tuzel · 2003
We present an automatic, real-time human tracking and observation system. Robustness and speed are the two major bottlenecks of the existing approaches. We improve upon the robustness and speed of the current state-of-art by integrating a mean-shift based model update technique with an adaptive change detection method. We also provide optimal solutions for several other stages including illumination compensation, skin color detection, shadow removal, morphological filtering, event analysis of a tracking system. In addition, we introduce a novel background refresh mechanism. Thus, the proposed framework is capable of handling shortcomings of template and correspondence based tracking approaches. The results with the ICVS-PETS data sets show the effectiveness of the algorithm.