A new algorithm of face detection based on differential images and PCA in color image

In this paper, a new algorithm of face detection based on differential images and PCA in color image is proposed which modified the eigenface technique. First, skin picking-up is carried out by color analysis of an image. Then, for finding the exact positions of faces, a finer matching is performed by eigenface in these detected fields and a mosaic template is used to suppress the false face. Considering the correlation between the pixels in a face image, the PCA method is performed both on the raw image and on its horizontal differential image and vertical differential image for feature extraction. And then the stability of the face detection is improved.