A COMPARATIVE STUDY OF CLASSIFICATION METHODS ON HUMAN SKIN DETECTION FROM RGB AND YCBCR REPRESENTED COLOR IMAGES

Skin detection has an important place in image processing. Human-computer interaction has made this study area very popular. The most common color space used in skin detection is Red Green and Blue but RGB space can be converted into YCbCr space. Both features give strong information about the properties of the images. In this study, RGB and YCbCr spaces are used to detect human skin. The extracted features are trained by several classification methods. The obtained features are used to segment the human skin by using the chosen classification algorithm and finally, the promising performance results are presented comparatively with the most commonly used classifications methods in the literature.