面向质量退化的车牌字符分割方法  

Character segmentation approach for poor license plate image

作　　者：徐云静[1] 卢书芳[1] 张元鸣[1] 高飞[1] XU Yun-jing;LU Shu-fang;ZHANG Yuan-ming;GAO Fei(College of Computer Science and Technology,Zhejiang University of Technology,Hangzhou 310023,China)

机构地区：[1]浙江工业大学计算机科学与技术学院

出　　处：《计算机工程与设计》2019年第11期3319-3324,3330,共7页Computer Engineering and Design

基　　金：国家自然科学基金项目(C12412135、61402410);浙江省重点研发计划基金项目(2018C01064)

摘　　要：为克服字符分割方法应对质量退化车牌时存在的鲁棒性差、效率低等不足,提出质量退化车牌字符分割方法。给出单行车牌结构特征的形式化表达;采用投影法实现字符粗分割,提出粘连字符的判断准则及去除粘连算法;给出单行车牌结构模式的定义及其求解算法,提出断裂字符和粘连字符恢复方法;给出误检和缺失字符位置推算方法。实验结果表明,该方法可有效处理因质量退化导致的车牌字符粘连、断裂、缺失等问题,在车牌字符分割精准度上有良好的表现。To overcome the disadvantages of poor robustness and low efficiency in dealing with quality degraded license plates,a license plate character segmentation method was proposed.A formalized expression of characteristic of single-line license plate was given.Characters were segmented roughly using projection method and adjoined characters were found and divided.Structure pattern of single-line license plate was defined and calculated,which was used to recover the fracture characters.A structure pattern-based prediction method was presented for misdetected positioning or lost characters.Experimental results show that the proposed method can effectively segment adjoined,fracture or lost characters in low quality license plate image.In addition,the proposed method shows good performance in segmentation accuracy.

关 键 词：车牌字符分割 车牌字符结构模式 粘连字符 断裂字符 投影法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]