An Algorithm for Human Face Detection in Color Image Based on Skin Color Segmentation

Face detection has taken more and more attention as a hot spot of study field which pattern ecognition and computer vision and Skin color has provided an important basis for people face detection. This paper presents novel EM algorithm for the face detection of image segmentation in the YUV color space. In order to separate the brightness information to chroma information in the expression of the colour, the algorithm uses the clustering method for skin color in colour spacer and improves the velocity performance of the regional fac separated from regional inhuman face. this paper also designs and optimized concrete steps. At last, colour images Segmentation experimental results and experimental analysis are given, and the papaer concludes the necessary conditions of the EM algorithm convergence.