摘要
为提高车牌定位准确率,提出一种改进的车牌定位方法。以小功率汽车牌照为研究对象,通过对HIS模型进行修正,分别得出包含车牌蓝色区域的二值图像和包含车牌字符的二值图像;将两幅二值图像合成一张图,利用车牌区域的几何特征实现对车牌区域的定位;用Hough变换和像素偏移法对倾斜的车牌进行矫正。实验结果表明,该方法与基于边缘检测的车牌定位方法相比,在耗时基本不变的情况下,图像定位准确率达到98%,倾斜矫正准确率达到99%,验证了该方法的有效性。
To improve the accuracy of vehicle license plate location,an improved method of license plate location was proposed.Taking license plates of small power cars as research objects,HIS model was modified to get two binary pictures that contained the blue region and the character region of a license plate respectively.The synthesis picture was composed by two pictures and the geometric features of the license plate region were used to locate license plates.The inclined plate was corrected using Hough transform and pixel offset method.Experimental results show that the image recognition accuracy rate increases to 98%compared to the method of license plate location based on edge detection,while the computation time of the two methods almost stays the same and the tilt correction accuracy rate increases to 99%.The results show that the method is efficient.
作者
张振威
苏燕辰
ZHANG Zhen-wei;SU Yan-chen(School of Mechanical Engineering,Southwest Jiaotong University,Chengdu 610031,China)
出处
《计算机工程与设计》
北大核心
2018年第11期3576-3583,共8页
Computer Engineering and Design
