基于颜色定位的车牌识别系统  

License Plate Recognition System Based on Color Positioning

作　　者：李鑫 黄杰 

机构地区：[1]武汉市黄陂区95903部队71分队

出　　处：《科技致富向导》2012年第32期209-210,共2页Guide of Sci-tech Magazine

摘　　要：车牌自动识别技术是智能交通系统中的一个重要方面。针对车牌定位技术中二值化中闽值的确定不稳定，本文提出基于颜色空间及灰度形态学的车牌定位方法，避开闽值确定的问题。该方法将采集的图像从RGB彩色空间转化到HSV彩色空间，再将图像转化为5级灰度图，然后采用灰度形态学运算并根据车牌特征从候选区域中准确地分割出原始图像中的车牌。在实践中取得了较好的效果，可供同类车牌识别系统参考。Automatic license plate recognition technology in the intelligent transportation system is an important aspect. License plate locatlon technology in the most unstable factor is a value of two in the determination of threshold, this paper based on the use of color space and gray-scale morphological approach for vehicle license plate locating, deliberately avoiding threshold determination problem. The main idea of this method is to collect the image in RGB color space conversion to HSV color space, and then converted to5 grayscale fmage, then uses the gray morphological operation and according to the characteristics of the license plate from the candidate area accurately segmented from the original image of the license plate. The method in practice had achieved good results, which can be used for reference of similar license plate recognition system.

关 键 词：颜色信息提取 车牌校正 擦除干扰区域 文字分割 模版匹配 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]