摘要
文章提出了一种基于纹理特征和颜色特征的车牌定位算法.算法首先利用车牌的纹理特征对车牌进行了初步定位,其中主要有图像预处理、边缘检测、数学形态学处理等步骤.最后利用车牌区域的颜色特征对车牌区域进行精确定位,主要包括空间转换、边界确定等步骤.对80张相片进行测试的结果表明,该方法准确率高、速度快.
A vehicle plate location method based on texture and color features is presented.The method locates primary vehicle plate area by making use of texture,including preprocessing image,detecting edge,mathematical morphology processing.At last,the method locates accurate vehicle plate area taking advantage of color features,including transforming color space,establishing verges.We have test 80 images and the experiments show that the method is of speediness and accuracy.
出处
《太原师范学院学报(自然科学版)》
2011年第1期78-80,110,共4页
Journal of Taiyuan Normal University:Natural Science Edition
基金
国家自然科学基金(60876077)
山西省自然科学基金(2009011018-3)
关键词
车牌定位
纹理特征
颜色特征
数学形态学
vehicle plate location
texture
color feature
mathematical morphology