Using self-organized and supervised learning neural networks in parallel for automatic target recognition

A hybrid approach to automatic target recognition (ATR) combining the complementary strengths of conventional image processing algorithms, artificial neural networks, and knowledge based expert systems is presented. The architecture employs parallel feature and pixel processing channels, the former using a self-organizing neural network and the latter using a supervised learning neural network. The feasibility of the hybrid automatic target recognition (ATR) approach to target detection, classification and recognition is demonstrated using (LADAR) data