The Study Of Human Face Detection Method Based On OpenCV

With the development of the intelligent information technology, identity recognition is highly demanded in the fields of national defense, security monitoring, long-range education and traffic control system. Human face, as the most important identity sign in the body, naturally becomes the detecting focus among other body features. Human face segmentation and detection, as the important project in the biometric identification, becomes the hot research focus of various major research institutions.Firstly, the development and application of the face segmentation and detection is summarized in this paper. Then, the basic theory of image segmentation is described in detail. Inspired by the image subdomain threshold segmentation technology, the moving foreground region binary image including human face is obtained by using the method of block adaptive threshold segmentation on the basis of background difference on the block background estimating method. Finally, moving foreground skin color region is detected by using the skin color model. By using face invariable feature discriminating method, the human face is quickly positioned.The major studies of this paper include:1. On the basis of image segmentation theory, the moving region foreground image segmentation of block adaptive threshold segmentation on the background image difference is designed and accomplished.Besides, the mathematical morphology is processed. 2. It proposes skin color region extraction and rapid face position by using skin color model and face invariable feature. In addition, it makes the comparison between AbaBoosting algorithm and the result of this experiment.The major works done in the research phase include: Human face detecting system is developed on the basis of OpenCV. The binary image and rapid face position is obtained by detecting the moving foreground region. It provides the platform for the real-time face detection in the video sequence.