Face detection based on SCNN and wavelet invariant moment in color image

The aim of this work is to describe a possible approach for the detection of face of arbitrary pose in color images, based on wavelet invariant moments and self-organizing competitive neural network. The method is capable of locating human faces over a broad range of views in color images with complex scenes. First of all, it uses the presence of skin-tone pixels to locate candidate face regions. And then a new method based on wavelet invariant moment and self-organizing competitive neural network is used to verify the candidate face regions. The experimental results show that the proposed algorithm has high speed and low error-detection rate, so it can be used in the real-time system. The main distinguishing contribution of this work is being able to detect faces at any degree of rotation in the image plane irrespective of their poses by using the wavelet invariant moments as input of the SCNN, whereas contemporary systems deal with upright, frontal faces. The other novel advantage of our method lies in its usage of self-organizing competitive neural network to detect face which greatly improve the efficiency of training procedure.