Face detection and tracking based on skin color and skin color moment
Yepeng Guan · Jisuanji gongcheng yu sheji · 2008
Novel and robust algorithm for face detection and tracking is presented.Moving regions are extracted based on background subtraction at first.Candidate face regions are extracted according to skin-color normalized RGB and HSV color model based on the ex- tracted motion region.Face information with certain size and some holes is used to verify the face candidates.The facial region is tracked and estimated based on skin color moment characteristics.The experimental results have shown that the proposed method not only attains high face detection rate,but also is robust to illumination and face pose variations.The processing speed is about 25f/s on an average across 480×360 images,which meets the demand of real time system for face detection and tracking.