Face Detection In Color Images Based On Skin Color And AdaBoost Algorithm

Human face detection is to locate the faces and determine their sizes in an image. Human face detection has recently been a hot spot of Computer Vision field. It is valuable in many different fields including security access control, visual surveillance, content-based information retrieval and so on.This paper is focus on complex background images. We build up a system for face detection based on skin color and AdaBoost algorithm, and do some testing in our face database.This paper firstly introduces the face detection's background, summarizes and analyzes current algorithms on face detection, then presents a new face detection method. The method is discussed on details as follows. Firstly, the method analyzes and compares complexion's clustering in different color space, and then establishes Gaussian Model based on the color space of YCgCr. After Skin color segmentation is carried out with the method of automatic threshold selection, the two-value images are obtained. Then we get candidate face regions using the filer based on mathematical morphology. After this, we can utilize some general knowledge of human face to select these candidate face regions. By doing these, the fake regions can be eliminated, and we can decrease those jobs in the further steps. Finally, we combine these candidate regions with the face regions detected by AdaBoost algorithm to improve the detection performance.At last, the method is implemented and experiment results are given out in this paper. At the same time, the paper points out the advantage and the weakness in our method and what we need to work hard in the future.