摘要
提出了利用小波分析、分形理论和神经网络技术进行车型图像的识别方法 ,主要包括用小波理论进行汽车图像的消噪和边缘检测、分形编码和神经网络分类等 3个部分 .经过实验证明 。
A method is put forward that can recognize the type of automobile, with adopts the technology of wavelet, fractal theory and neural networks. The paper is made up of three parts: first is to reduce the image'noise and check the edge of automobile, second is fractal coding, last is to classify the automobile images using neural networks. Proved by experiences, the method is better than others.
