Automatic target recognition (ATR) is an important issue in the military field, the topic of the ATR system is the pattern recognition and classification. In the paper, we present an approach for building an ATR system with improved artificial neural network to recognize and classify the typical targets in the army field. The invariant features of Hu invariant moments and roundness were selected to be the input of the neural network for they have the invariance of rotation, translation and scaling. The pictures of the targets are generated by the 3-D models to improve the recognition rate for it is necessary to provide enough pictures for training the artificial neural network. The simulations prove that the approach can implement the task of ATR system in high recognition rate and real time.