Human Face Detection in Color Images

Abstract

Color characteristics are proposed to be utilized to detect human faces in color images. The proposed face detection method first uses a neural network to segment out the candidate face regions. Then, and energy thresholding method that can take the shape, the color, and the edge characteristics of the face features into the extraction process is devised to extract the lip. Finally, three shape descriptors of the lip feature are used to further verify the existence of the face in the candidate face regions. Our experimental results show that this method can detect faces in the images from different sources in an accurate and efficient manner.