Gabor filtering usinc cellular neural networks and its application to face/object recognition

Gabor filtering is useful for intelligent image processing, but it requires huge computational power. We propose a new Gabor filtering algorithm using a discrete-time cellular neural network (CNN) circuit, which is suitable for pixel-parallel LSI implementation. The proposed algorithm utilizes transient states of the CNN to obtain Gabor coefficients. We apply this Gabor filtering algorithm to face/object recognition based on dynamic-link matching. A PC system for recognition of natural scene images with human faces and various objects has been constructed. The system also includes FPGA implementation of nonlinear resistive networks.