基于模板匹配的发票号码识别算法  

Identification algorithm for invoice number based on template matching

作　　者：宫义山[1] 王鹏[1] 

机构地区：[1]沈阳工业大学信息科学与工程学院,沈阳110870

出　　处：《沈阳工业大学学报》2015年第6期673-678,共6页Journal of Shenyang University of Technology

基　　金：辽宁省教育厅科技计划项目(201344077)

摘　　要：针对数字识别系统中的发票号码识别问题,提出了一种有效的基于模板匹配的发票号码识别算法.通过对发票数字图像进行前期处理以获得单个字符图像,根据印刷体数字二值化后矩阵特点,基于模板匹配的思想,将待识别字符图像与模板图像进行信息区叠加,对叠加后的图像进行不匹配像素统计,从而得到匹配系数,取匹配系数最小值为最佳匹配.结果表明,该方法简单有效,抗干扰性强且识别率较高,识别发票图像准确率可达99%.In order to solve the identification problem of invoice number, an effective identification algorithm for invoice number based on template matching was proposed. Through the pretreatment of invoice number image, the single character image was obtained. According to the characteristics of the matrix after the binaryzation of printed number and based on the idea of template matching, the information area superposition between the character image to be identified and the template image was performed. In addition, the unmatched pixel statistics for the image after the superposition was carried out, and thus the matching factor was obtained. The minimum matching factor was regarded as the optimum matching. The results show that the proposed method is simple and effective, and has strong anti-interference performance and high identification rate. Moreover, the identification accuracy of invoice image can reach 99%.

关 键 词：发票号码 模板匹配 单个字符图像 不匹配像素 二值化 矩阵 信息区 匹配系数 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]