Real-time Face Detection For Video Surveillance

In recent years, video surveillance in airports, financial, security and other fieldshas been extensively applied; however, existing system has some defects. The rapiddetecti6n and segmentation algorithm for faces on video surveillance which has beenpresented by this paper, can effectively improve the efficiency and intelligence onthis.Face detection problem has been researched for more than 20 years, butso far,due to the complexity of face detection problem, while making many researchers havedevoted to this problem for many years, this problem is still not fully solved. Theauthor discussed this problem with scientific research on face detection issues.Working mainly focus on the following three aspects:1. Improved the skin color detection algorithm based on YCbCr color space. Anuse of the previous frame skin color detection result on video sequence and colorknowledge with simplified rules of color judgment, and the use of the improvedmedian filtering method on binary image simplifies the skin color detection algorithm.Achieved a rapid detection of skin regional on color video frames algorithm based onthe YCbCr color space.2. Use the binary skin color template matching to achieve the face detectionmethod basis on skin color segment. Achieved a real-time face detection algorithm onvideo surveillance.3. Use OpenCV with AdaBoost based to detect face and applied it to videosurveillance, Achieved a real-time face detection algorithm for video surveillance.