车牌定位与字符分割的新方法
|
摘要
车牌定位与车牌字符分割是车牌识别系统中的2个关键技术,在车牌定位过程中,采用HSV颜色分割算法和投影法相结合的方法,对现有的算法进行了改进。在字符分割过程通过实验确定出最佳二值化阈值,对图像进行二值化处理,通过垂直投影实现字符的有效分割。实验结果表明,使用新方法,车牌识别率和识别的鲁棒性有显著提高。
The license plate location and license plate character segmentation are two key technologies in the license plate recognition system. In the process of license plate location, this paper uses HSV color segmentation algorithm and projection method to combine and improve the existing color segmentation algorithm and projection algorithm. In the process of character segmentation, the best two thresholds are determined through experiments, and the two values of images are processed, and then the effective segmentation of characters is achieved through vertical projection. The experimental results show that the robustness of license plate recognition rate and recognition is greatly improved by using the new method.
The license plate location and license plate character segmentation are two key technologies in the license plate recognition system. In the process of license plate location, this paper uses HSV color segmentation algorithm and projection method to combine and improve the existing color segmentation algorithm and projection algorithm. In the process of character segmentation, the best two thresholds are determined through experiments, and the two values of images are processed, and then the effective segmentation of characters is achieved through vertical projection. The experimental results show that the robustness of license plate recognition rate and recognition is greatly improved by using the new method.
引文
[1]刘万军,姜庆玲,张闯.基于CNN彩色图像边缘检测的车牌定位方法[J].自动化学报,2009,35(12):1503-1512.
[2]王义兴,黄凤岗,韩金玉,等.基于颜色搭配与纹理特征的车牌定位方法[J].中国图象图形学报,2009,14(2):303-308.
[3]张吉斌.基于图像处理及支持向量机的车牌识别技术研究[D].兰州:兰州交通大学,2013.
[4]黄灿.基于局部特征的车辆识别[D].上海:上海交通大学,2010.
[5]杜晓刚,王建中,白艳萍,等.一种快速有效的混合倾斜车牌校正方法[J].数学的实践与认识,2012,42(19):87-95.
[6]马宁,李斌.基于神经网络集成的车牌字符识别[J].安徽广播电视大学学报,2012(2):116-120.
[7]SHAN Baoming.License Plate Character Segmentation And Recognition based on RBF Neural Network[C]//ETCS 2010,the 2nd International Workshop on Education Technology and Computer Science,6-7 March,2010,Wuhan,China,2010:86-89.
[2]王义兴,黄凤岗,韩金玉,等.基于颜色搭配与纹理特征的车牌定位方法[J].中国图象图形学报,2009,14(2):303-308.
[3]张吉斌.基于图像处理及支持向量机的车牌识别技术研究[D].兰州:兰州交通大学,2013.
[4]黄灿.基于局部特征的车辆识别[D].上海:上海交通大学,2010.
[5]杜晓刚,王建中,白艳萍,等.一种快速有效的混合倾斜车牌校正方法[J].数学的实践与认识,2012,42(19):87-95.
[6]马宁,李斌.基于神经网络集成的车牌字符识别[J].安徽广播电视大学学报,2012(2):116-120.
[7]SHAN Baoming.License Plate Character Segmentation And Recognition based on RBF Neural Network[C]//ETCS 2010,the 2nd International Workshop on Education Technology and Computer Science,6-7 March,2010,Wuhan,China,2010:86-89.