The Study Of Algorithms About Face Detection Based On Complexion

Face recognition technology, as a major field of pattern recognition and computer vision research, is particularly significant to applications such as identity authentication, video monitoring, and frontier defense security etc. Two key techniques accounts the most for a human-face recognition system: one is face detection and orientation; the other is feature abstraction and recognition from unified human-face image.Being a primary phase of face recognition, vision point tracing, and face image compression etc., inspecting human face out from the image background is the very first step. However, the result of face detection is usually affected by the background, brightness changing or head posture of the image and so on, which makes the process of detection more complicated. The paper summarizes concepts, applications, key technical problems, and latest research outcomes for the face detection technology, and makes introductions to as well as comparisons between some successful systems applied in face detection field. In addition, the thesis brings out a complexion-based algorithm for the chromatic-imaged face detection. This approach firstly conducts a dimension unification operation uponthe input image followed by necessary light compensation, then conducts a binary-value transformation operation upon the image with dual threshold using both YCbCr color space and RGB color space; following that the system further eliminates noise by means of open operations or closed operations based on complexion partitioning, and as the final step, the system judges from these complexion areas according to factors like scale, shape, hair, filling ratio etc., and then discards false ones, thus identifies human face areas.The results of experiments show that this detection approach features a rapid inspecting speed and is insensitive to the face posture; meanwhile, since it uses an area partitioning method, effective multi-face discrimination is easy to achieve in application, as shown by experimental results.