ARTMAP与模糊逻辑结合的字符智能识别方法  

CHARACTER RECOGNITION WITH THE COMBINATION OF ARTMAPAND FUZZY RULES

作　　者：徐立亚[1] 戚飞虎[1] 蒋隽 林纯青[1] 

机构地区：[1]上海交通大学计算机科学与工程系

出　　处：《红外与毫米波学报》1998年第5期363-368,共6页Journal of Infrared and Millimeter Waves

基　　金：国家自然科学基金;国防预研基金

摘　　要：研究了将ＡＲＴＭＡＰ神经网络与模糊规则相结合的字符识别方法．ＡＲＴＭＡＰ用于对字符的学习、训练，而模糊规则和ＡＲＴＭＡＰ共同用于对字符的智能识别．实验证明这种将模糊技术与神经网络相结合的混合系统具有较高的识别率和较快的识别速度，采用ＡＲＴＭＡＰ神经网络只需要较少的网络训练时间．A method for character recognition was presented, which uses ARTMAP and fuzzy rules. The neural network generates weights, which can be used to determine fuzzy regions. The fuzzy rules recognize most of the test samples and ARTMAP recognizes the rest of characters. The experiment result proves that the combining method has higher ratio for character recognition. ARTMAP is a fast and efficient neural network, which has no local minimum point.

关 键 词：神经网络 ARTMAP 模糊规则 字符识别 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP391.41[自动化与计算机技术—控制科学与工程]