Human face recognition is an important subject in the area of pattern recognition, which has a wide range of potential applications. This paper did some searching in the range of face detection and recognition. It include Canny operator, Hough transform, PCA , mathematical morphology, face color model and face features extraction. Based on the old Canny operator, we proposed a new edges extraction method. Based on the result of line detection, we proposed a new disk/circle detection algorithm through Hough transform. We also proposed a new PCA method combined with edges, Improved the toss curve method extracting chin profile and tried to use Sin curve to extract chin profile. In the last, we achieved to constitute a system used to detect and recognize human face. |