A method of skin color identification based on color classification

This paper describes a method to identify human skin region in an image on a color classification result. The method classifies the colors of all pixels in the image into several classes through K-means algorithm and segments the image into several parts according to the color class that each pixel belongs. Each class of color is represented by a color feature vector. Find the class whose feature vector has the minimum distance to the skin color feature vector previously defined in the color space. Take this class as human-skin color in this image and the skin color region can be identified. This method does not need an accurate skin color model which can only be obtained through resource-costing training processes. It offers a relatively simple way to find the regions that is most similar to skin color and so to identify the regions as skin regions. Experiment results are shown to prove the effectiveness.