基于模板匹配的车牌字符识别算法研究  

Research on License Plate Character Recognition Algorithm based on Template Matching

作　　者：王帅 刘光宇 李俊松[1] 程远 周豹 赵恩铭 张令通 WANG Shuai;LIU Guang-yu;LI Jun-song;CHENG Yuan;ZHOU Bao;ZHAO En-ming;ZHANG Ling-tong(Dali University,Dali 671003,China;Shanghai Jiao Tong University,Shanghai 200030,China;The People's Liberation Army,Dali 671003,China)

机构地区：[1]大理大学,云南大理671003 [2]上海交通大学,上海200030 [3]中国人民解放军32268部队,云南大理671003

出　　处：《武汉船舶职业技术学院学报》2024年第1期97-102,共6页Journal of Wuhan Institute of Shipbuilding Technology

基　　金：国家自然科学基金资助项目:“基于自旋交换泵浦技术的高透过率、超窄带原子滤光器的研究”(项目编号:62065001);云南省地方本科高校基础研究联合专项资金项目“用界面追踪数值模拟方法研究微纳尺度粗糙结构对材料表面润湿性的影响”(项目编号:202101BA070001-054);海洋智能装备与系统教育部重点实验室开放基金项目“基于深度CNN的海底废弃物识别技术”(项目编号:MIES-2023-02)。

摘　　要：基于模板匹配的字符识别是车牌识别中重要的研究方法,该方法下组成的车牌识别系统涵盖了图像预处理、车牌定位、字符分割以及模板匹配等主要内容。该算法首先对拍照获取的车牌图像进行灰度化、边缘检测、阈值分割等预处理,其目的是为得到只含有车牌信息的图像;然后将车牌图像进行腐蚀处理,确定车牌区域,并且通过二值化得到突出车牌字符信息的二值图像,以便于进行字符分割;最后再将分割出的字符与字符模板进行匹配得到准确的识别结果。实验结果表明:基于该方法所实现的车牌识别系统,其识别过程简单且便捷,对于车牌的定位较为精准,能够较好地满足车牌识别的基本要求。Template matching-based character recognition is an important research method in license plate recognition,under this method,the composed license plate recognition system covers main components such as image pre-processing,license plate localization,character segmentation,and template matching.Then the license plate image is etched to determine the license plate area,and the binary image highlighting the license plat character informa-tion is obtained through image binaryzation,so as to facilitate the character segmentation.Finally,the segmented characters are matched with the character template to obtain accurate recognition results.The experimental results show that the recognition process of the license plate recognition system based on this method is simple and convenient,the license plate positioning is more accurate,and can better meet the basic requirements of license plate recognition.

关 键 词：车牌识别 图像处理 车牌定位 模板匹配 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]