R Research On Face Detection In Color Images

Human face detection plays an important roll in applications such as video surveillance, human computer interface, face recognition, and face image database management. Face detection and recognition technology is one of the hottest spots in pattern recognition, computer vision and image understanding system; it has integrated multi-disciplinary knowledge such as image understanding, computer vision, databases and artificial intelligence, and with the rapid development it get more and more attention and has made many valuable research results. But now the face detection system can not reach our expectation, especially with the bad environment.A face detection algorithm for color images in the presence of varying lighting conditions as well as complex backgrounds is proposed by us. Our light compensation method which is based on different color space from-wide-to-slender and from-whole-to-part instead of estimating the surround light conditions is brought out. The method does not have to estimate the direction, strength and the nature of light source (such as: whether it is point light source, parallel light, single-source or multi-source, etc.) The method with different light compensation strategy adapts different light circumstances well. With the description of skin color information based on mixed Gaussian model and the Pixel Pitch Center Ma-distance to calculate the color similarity, the algorithm can gain a good result. Based on a novel lighting compensation technique and a nonlinear color transformation, our method detects skin regions over the entire image and then generates face candidates based on the spatial arrangement of these skin patches. It can classify the face area and non-face area effectually.It validate that the problem which is caused by light condition can be solved by the method with both the theory and experimentation. it can enhance the accuracy of division between regional and non-skin color regional so that the face detection performance in different light conditions has been greatly improved .This method also has high efficiency operation and robustness,etc.