Face recognition system using skin detection in RGB and YCbCr color space

In this paper, we present a face recognition system using skin segmentation as feature reduction. Human Skin detection reduces the face region research in the image. In addition, color skin is a robust information face to rotation; scale and illumination variation. Skin regions are extracted using a set of bounding rules based on the skin color distribution obtained from a training set. The proposed RGB-YCbCr skin color model for skin region segmentation was evaluated on a set of three databases. The set was choice in a controlled and uncontrolled environment. Two classifiers are studied to build our face recognition system; the K nearest neighbor applied for each channel of the segmented image, whereas the second used the Principal component analysis PCA as reduction method followed by knn classifier. Best recognition rate were obtained for the first classifier.