在线字轮数字式仪表的自动读取技术  

Automatic recognition technology for online word wheel's digital meters

作　　者：王荣[1] 胡晓兵[1] 王勇[1] 李维[1] 

机构地区：[1]四川大学制造科学与工程学院,成都610065

出　　处：《计算机应用》2008年第S2期354-356,共3页journal of Computer Applications

摘　　要：针对字轮数字式仪表传统手工读数所带来的费时、费力、时效性低等问题,提出了一种模板匹配与特征提取相结合的改进数字识别方法。其基于数字的多个明显形状特征,并给予不同加权值的模式匹配法有效地改进了识别正确率。实验证明,该读取技术具有较高的实时性与鲁棒性。An improved pattern matching numeric identification methods was introduced,which based on several shape features and then gives these features different weighted value to get a final template matching result.This method effectively improves the identification rate and solves the problems of time-consuming,laborious,low efficiency in the conditional manual word wheel's digital meter reading work.The experiments show that this system has good characteristics of real-time and robustness.

关 键 词：在线 模式识别 形状特征 加权值 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]