Gabor-type image filtering with cellular neural networks

Gabor filters have been used as preprocessing stages in several different types of image processing and computer vision applications. One drawback is that they are computationally intensive on a digital computer. Here, we describe the theory underlying a cellular neural network architecture which simultaneously computes the outputs of two filters similar to odd and even phase Gabor filters. By computing the filter outputs with less power and in less time than required by serial digital computers, an analog VLSI implementation of this CNN could relieve the computational bottleneck associated with Gabor filtering image processing algorithms.