基于轮廓识别和BGR颜色空间的车牌定位  

License Plate Location Based on Contour Recognition and BGR Color Space

作　　者：漆世钱 QI Shi-qian(Department of Electronic Technology,China Coast Guard Academy,Ningbo 315801,China)

机构地区：[1]武警海警学院电子技术系,浙江宁波315801

出　　处：《计算机技术与发展》2020年第12期176-180,共5页Computer Technology and Development

基　　金：国家自然科学基金(61401105);武警海警学院2019年实战化教学改革项目(YB2019013)。

摘　　要：车牌定位技术是车牌识别系统的关键步骤,车牌定位的效果会影响后面对车牌字符分割和识别的成功率,进一步影响整个车牌识别系统的成功率和识别速度。为解决车牌在角度偏移、光照不均匀的影响下定位难、定位慢的问题,提出一种基于轮廓识别和BGR颜色空间的车牌定位方法。将车牌定位分为粗定位和精细定位两部分,粗定位利用轮廓识别技术找出车牌的大致区域,基本步骤包括轮廓检测、矩形区域识别、矩形筛选和截取;在精细定位中,对图像逐像素点遍历,读取各点的B、G、R值,通过BGR处理提取车牌关键位置,然后校正和截取可能倾斜的车牌图像。实验结果表明,在常规环境中,该方法对车牌的定位准确度高达96.5%,算法运行时间低至612 ms,达到了预期效果。License plate location technology is the key step of license plate recognition system.The effect of license plate location will affect the success rate of license plate character segmentation and recognition,and further affect the success rate and recognition speed of the whole license plate recognition system.In order to solve the problem of difficult location and slow location of license plate under the influence of angle offset and uneven illumination,a license plate location method based on contour recognition and BGR color space is proposed.The license plate location is divided into two parts:rough location and fine location.Rough location uses contour recognition technology to find out the approximate area of license plate.The basic steps include contour detection,rectangular area recognition,rectangular filtering and interception.In fine location,the image is traversed pixel by pixel,the B,G and R values of each point are read,the key position of license plate is extracted through BGR processing,and then the possibly tilted license plate image is corrected and intercepted.The experiment shows that in the conventional environment,the accuracy of the proposed method is as high as 96.5%,and the running time of the algorithm is as low as 612 ms.

关 键 词：车牌定位 轮廓识别 BGR处理 矩形筛选 角度倾斜 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]