A head pose detection algorithm based on template match

Visual analysis of human movement is a hot topic in the field of computer vision. A novel algorithm- column integral, speed up the head location, and has the more precision than previous algorithm. A head turning detection algorithm has been presented, which based on the color information of each space branch area and template match, and it is simple and effective. The experiment result indicate that the proposed algorithm has a good real-time performance, the improved head positioning algorithm can reduce by 1.24 ms than the similar algorithm in once position, and accuracy of the head turning detection algorithm can reach 92%.