摘要
改变传统的手写体汉字识别的分类决策过程,从仿生模式识别的认知观点出发,借鉴了人类在四角号码字典中查字的方式.从手写体汉字图像提取8种典型形态特征进行编码,构造形态编码字典.从而将汉字识别转化为模仿人们在八形态编码字典中查字的过程.仿真实验结果表明,这种采用八形态编码的手写体汉字识别方法是可行的、有效的.
Based on the cognitive view of the bionic pattern recognition, changing classification decision process of the traditional handwritten Chinese characters recognition, and drawing lessons from the way of human's consulting a four- comer coding dictionary, the author extracts eight shape features from handwritten Chinese characters image to encode, structures shape coding dictionary, and converts Chinese characters recognition into the process of looking up the words in eight - shape coding dictionary by modeling a person. The simulation results show that the Chinese characters recognition method is feasible and effective.
出处
《淮北师范大学学报(自然科学版)》
CAS
2012年第2期38-42,共5页
Journal of Huaibei Normal University:Natural Sciences
关键词
仿生模式识别
四角形态
笔划形式
字元
编码
bionic pattern recognition
four - comer shape
strokes form
character
coding