基于HALCON的喷码字符识别技术的研究与实现  

Research and implementation of ink-jetted code character recognition technology based on HALCON

作　　者：肖海俊[1,2] 葛广英[1,2] 姚坤[1,2] 尹红敏[1,2] 

机构地区：[1]聊城大学物理科学与信息工程学院,山东聊城252059 [2]山东省光通信科学与技术重点实验室,山东聊城252059

出　　处：《现代电子技术》2015年第15期95-98,共4页Modern Electronics Technique

基　　金：国家自然科学基金项目(10874063)

摘　　要：为了实现产品包装上喷码字符高效、快速、准确地识别,设计了一种基于HALCON的喷码字符识别系统。首先采用四次不同结构元素的闭运算准确提取字符,然后再采用BP网络训练样本,根据字符的特征对样本进行分类,生成相应的字符库,并调用该字符库读取字符,进而识别出字符符号。实验结果表明,这种方法运算速度快、准确性高。To realize efficient,rapid and accurate ink?jetted code character recognition on product packaging,an ink?jetted code character recognition system based on HALCON was designed. The closed operation of four different structural elements is adopted to extract the character. The samples are trained by using BP network,and classified on the basis of character features. Then the relevant character library is generated. The character is read by calling the character library to recognize character sym?bols. The experimental results show that this method has fast operation speed and high accuracy.

关 键 词：喷码字符 HALCON 闭运算 BP网络 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]