A New Method of Single Color Face Detection Based on Skin Model and Gaussian Distribution

The speed and the accuracy of face detection are both very important for network sensors applied in human-computer interface. A new method of face detection, which is based on skin color clustering features and the characteristic that the projection of facial pixels onto the Euclidean Space with two dimensions is similar to the Gaussian distribution, is presented in this paper. Firstly, this approach utilizes the skin color features to transform the original image into a binary-image. Then, two curves are obtained by projecting the binary-image to X and Y axes respectively, from which Gaussian curves are calculated. Finally, an accurate face region is found through the solution of Gaussian equation. Simulation shows that the proposed method is fast in detection and needs less computation