Face Detection And Its Application In Face Recognition

In recent years, face detection has attracted the attention of many researchers, as aprimary and key step for video surveillance, human-computer interaction, face recognitionand face image database management. The purpose of face detection is to realize rapiddetections from complex background images. Skin color is important information incolorful face images. It does not depend on the specific features of facial and is relativelystable. Although different people’s skin color vary widely, different color of skin on thechromaticity tend to be similar, they are clustered in the shade. So the color feature canbe used to quickly extract the target area in the complex images.Eyes detection and lips detections are the key points of face recognition. Mouth as animportant organ, has its unique characteristics in color and shape, the information in lips isnot easy to hide, thus it can be used to extract features. Lip detection especially under thebackground of complex lips, has great development potential in public securitymonitoring.This paper firstly analyzes and compares among several color spaces that havingbeen commonly used to select preferable YCbCr color space to build skin color models, tomake similarity segmentation and to do binarization processing. After all candidates areextracted, we combine the geometric feature of face and a prior knowledge to detect theface region. Secondly, this paper extracts the most important organs of human face,namely eyes and mouth. Finally, the extracted eye region and mouth region are applied ina expression recognition system, and their contribution to expression recognition accuracyis compared and analysed.