A Lighting Insensitive Face Detection Method on Color Images

In this paper we proposed a lighting insensitive face detection method depending upon the skin color and edge of the input color image. First, image enhancement is performed, especially if the image is acquired from an unconstrained illumination condition. Next, the skin segmentation in YCbCr and RGB space is conducted and the result is combined with edges of the input image. Candidate verification using primitive shape features of face is applied to decide which of the candidate regions corresponds to a face. The proposed face detection method is invariant to size, pose, expression and illumination conditions.