29P-C-2 粒子フィルタと窓を使用したジェスチャ追跡(トラッキング・最適化,一般講演)

抄録

The human gesture tracking is needed in many applications, e.g., the man-machine interface, recognition of the hand language, the surveillance cameras, and etc. Recently, the advent of Microsoft's "Kinect" makes the gesture tracking quite easy, if it is used in a restricted environment like indoor However there are still large demands of gesture tracking using only video cameras. In this paper, we propose a method which tracks the human gesture thorough video images using the particle filter This method is based on the inspection that a sequence of human gestures can be divided into basic patterns of human motions. We examine the effectiveness of the method by experiments in which gesture tracking is done for the radio gymnastic exercises.