基于角点检测和边缘提取的车牌定位方法  

Vehicle License Plate Locating Algorithm Based on Corner Detection and Edge Extraction

作　　者：徐志刚[1] 朱红蕾[1] 

机构地区：[1]兰州理工大学计算机与通信学院,兰州730050

出　　处：《科学技术与工程》2008年第14期3985-3988,共4页Science Technology and Engineering

基　　金：国家科技支撑计划(2006BAF01A21)资助

摘　　要：车牌定位技术是实现车牌自动识别的关键技术之一。根据车牌特征,设计了一种基于角点检测和边缘提取的车牌定位方法。该方法首先对原始图像进行阈值分割,其次采用数学形态学方法增强车牌边缘,接着利用角点检测方法对车牌进行粗定位,最后通过边缘提取来实现车牌图像的精定位。通过实验分析,证明该方法定位效果较好。Vehicle license plate locating technology is one of the key technologies in automatic LPR （license plate recognition ）. Aiming at the license plate features, an algorithm based on corner detection and edge extraction is introduced to locate license plate of vehicle image. Thresholding is adopted at first. Second, the mathematical morphology method is used to enhance plate edge. Then, the corner detection method is applied to rough locating of the license plate. At last, the refined locating is achieved by the edge extraction method. Experimental resuits demonstrate its effectiveness and efficiency.

关 键 词：车牌定位 阈值分割 数学形态学 角点检测 边缘提取 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]