A simple neural network approach to invariant image recognition

In this paper we propose a simple neural network architecture for invariant image recognition. The proposed neural network architecture contains three specialized modules. The neurons from the first module are connected in a cellular neural network structure, which is responsible for image processin g: edge detection and segmentation. The second module is a feed forward neu ral network for invariant feature extraction from the sensorial layer: comp utation of the pair distribution function and bond angle distribution funct ion. The third module is responsible for image classification. An applicati on to the face recognition problem is also presented.