基于特征加权模糊模板匹配的字符识别  

Character recognition based on feature weight fuzzy template matching

作　　者：朱颢东[1] 李红婵[1] 

机构地区：[1]郑州轻工业学院计算机与通信工程学院,河南郑州450002

出　　处：《兰州理工大学学报》2013年第1期68-71,共4页Journal of Lanzhou University of Technology

基　　金：河南省基础与前沿技术研究计划项目(102300410266;122300410287);郑州轻工业学院博士科研基金(2010BSJJ038)的资助

摘　　要：字符识别算法能够直接影响字符识别的精度和速度.把特征加权模板匹配算法和基于模糊原理的判别准则结合起来,设计一个字符识别算法.该算法首先对标准模板匹配算法进行改进,然后在此基础上以模糊理论作为判断标准来进行字符识别.从理论分析和实验结果这两方面来看,所提字符识别算法能够充分区分开字符笔画和非笔画部分对字符识别的影响大小,可以有效提升识别率和鲁棒性,性能较优.The accuracy and rate of character recognition will be affected directly by the character recognition algorithm.By combining feature weight template matching algorithm with fuzzy identification criterion,a character recognition algorithm was designed in this paper.The proposed algorithm firstly improved the standard template matching algorithm and then employed fuzzy theory as identification criterion to recognize character.The theoretical analysis and experimental results showed that the proposed algorithm could be used to separate completely the influences of character stroke and character tailing on character recognition,so that the recognition rate and robustness were effectively improved,resulting in more optimal performance.

关 键 词：字符识别 特征加权 模板匹配 模糊理论 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]