摘要
提出了一种基于OpenCV的车牌识别和校正方法。该方法采用纹理特征和颜色信息相结合的算法对车牌进行定位,采用空间域分析的方法对倾斜车牌进行校正。在Visual Studio 2010和OpenCV2.4.1平台下验证,该方法能够对不同场景下不同种类的车牌进行定位和校正。该方法有较高的定位和校正准确率,并且能够应用于多种场合。
Vehicle license plate recognition is an important technology in traffic management. Location and correction of vehicle license plate is the most important part of vehicle license plate recognition. This paper presented a method of vehicle license plate location and correction based on OpenCV. The proposed method got the location of plate using both color and texture information. And it used space domain analysis to correct plates. Experiments using Visual Studio 2010 and OpenCV 2.4.1 confirmed it's effectiveness in different types of plates. And it can be used under different conditions.
出处
《计算机应用》
CSCD
北大核心
2013年第A01期223-226,235,共5页
journal of Computer Applications
基金
浙江大学CAD&CG国家重点实验室开放基金课题(A1213)
华东理工大学国家大学生创新性实验计划项目(S11075)
关键词
OPENCV
车牌
识别
定位
校正
纹理特征
颜色特征
OpenCV
vehicle license plate
recognition
location
correction
texture feature
color feature