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

摘要

Face detection has attracted increased attention as a hot research topic in the field of pattern recognition and computer vision. Skin color has provided an important basis for human face detection. This paper presents a novel expectation-maximization (EM) algorithm for the face detection of image segmentation in the YUV color space. To separate brightness information from chroma information in the expression of color, the algorithm uses the clustering method for skin color in the color space and improves the velocity performance in the separation of a regional face from a regional nonhuman face. This paper also designs and optimizes the concrete steps. Finally, the segmentation experimental results and experimental analysis in terms of color images are given; in addition, the paper provides the necessary conditions for EM algorithm convergence.