Tracking of Human Gestures in RGB Image Sequence Based on Statistical Inference and Three-Dimensional Human Model Taken with Ocellar CCD Camera

摘要

Many methods for tracking of Three-Dimensional human gestures have been proposed by using multiview scheme. However, at the present state these methods have been very far from the stage of practical application due to its high-cost. This paper describes a new method of Three-Dimensional human gesture from image sequence taken with Ocellar CCD camera. This tracking system is composed based on statistical inference and Three-Dimensional human model, and the Occlusion problem is solved by bottom-up and top-down approachs. We demonstarated the tracking system of human gestures.