An Improved AdaBoost Algorithm for Face Detection

Face detection is widely used in human-computer interface, security access, visual surveillance and other fields, but the accuracy and speed are two key problems which need to be solved. We proposes an improved face detection method based on the existing AdaBoost algorithm. By locating the face area according to the structural features and skin color features, we can judge whether it is the final target face according to the overlap of the detected target area. Experiments show that the method has obvious effect in improving the detection accuracy.