Human facial feature extration is the key of facial image analysis, which has been widely applied to face recognition, 3D facial reconstruction and face image compression. This paper focuses on automatic extraction of feature points from the two-dimensional digital face photo. Some improvements and simplifications are made on the basis of the previous research. The automatic extraction process is adjusted and gray is before data processing. The face regional calibration method is the edge detection and binaryzation after clustering and the result is under control of the integral projection combined with the physiological structure. Experimental results show the method can effectively extract 2D face image feature and be practical.
展开▼