Skin Recognition Based on Color Image Enhance Algorithm

Traditional skin recognition process include parameters ensuring and recognition. Most recognition focused on the accurate of parameters and taked out method of filter light information of image. Experiment proved that the light imformation can not be filter directly from image. This paper take out algorithm of skin recognition based on image enhance method. Firstly, image is enhanced non-linear transform to enhance the dark pixels. Then adjust the contrast with diagram banlance and local enhance method. In the end, renew the image with linear transform. After this process, we recognize the image with ellipse clusterring skin color model. This method has the advantage of high recogniton rate and strong robust.