摘要
人工神经网络是信息科学、脑科学、神经心理学等诸多学科近年来共同关注的研究热点。由于神经网络具有良好的抽象分类特性,使其成为解决图像识别相关问题的有效工具。在简述图像识别过程的基础上,重点讨论利用BP神经网络对图像进行识别,用MATLAB完成对神经网络的训练和测试,获得比较满意的结果。
Artifical neural network is the hot researching subject that is concemed commonly by a number of scientific areas such as information science, brain science and neuropsychology etc. over the recent years. Because of the excellent characteristic of neural network in abstract classification, it becomes the effective tool in solving the relevant problems in image recognition. During introducing the process of the image recognition,discussion is mainly focused on recognizing the image by using BP neural network and on accomplishing exercise and test by MATLAB. The result is comparatively satisfactory.
出处
《机械制造与自动化》
2009年第2期78-80,共3页
Machine Building & Automation
关键词
反向传播网络
数字识别
图像处理
BP network
numerical recognition
image processing
