Frontal Face Detection Based on Skin Color Segmentation and Partial Face Classifier

This paper proposes a frontal face detection system based on skin color segmentation and partial face classifier. Firstly, the non-skin color pixels in the input image are removed by the skin color classifier which uses the mixed color model in RGB and CbCr spaces. Secondly, possible face block positions are detected by using the face detector which consists of the partial face classifier and the facial structure test. The experimental result for the proposed face detection system showed 95.3% detection rate.