提出了一种利用前向三层神经网络及计算机视觉技术对汽车车型进行自动识别的方法，包括汽车车型图象的分割，特征提取以及基于前向三层神经网络的自动识别和分类，文中的所有算法均在586/133微机上实现，识别和分类效果良好。

This paper presents a method using feed-forward three layers neural networks and computer vision technology to recognize automobile models. It includes automobile model image segmentation, character abstraction, automatic recognition and classification based on feed-forward three layers neural networks. All algorithms in this paper are verified at 586/ 133 computer and experimental results are efficient for recognition and classification.