用改进的投影法实现车牌字符分割  

License plate character segmentation realization based on improved projection algorithm

作　　者：杨峰[1] 潘越[1] 李建奇[1] 

机构地区：[1]湖南文理学院电气与信息工程学院,湖南常德415000

出　　处：《湖南文理学院学报（自然科学版）》2017年第4期34-35,39,共3页Journal of Hunan University of Arts and Science(Science and Technology)

基　　金：国家自然科学基金(61403136);湖南省自然科学基金(14JJ5008)

摘　　要：提出了一种改进的车牌字符分割投影方法。首先用Otsu算法对车牌图像二值化,再对边框、铆钉进行处理,消除其带来的不利影响,最后利用Hough变换做直线检测,确定车牌的倾角后进行矫正。各环节均在Matlab环境下完成。试验结果表明,在选定的使用未经改进的直接投影无法分割的图像样本中,94%的样本通过改进的投影法被正确分割。An improved license plate character segmentation projection algorithm is introduced. First, image binarization is based on Otsu algorithm. Second, unfavorable effects induced by plate edge and rivet are overcome by means of any treatments. Third, line detecting based on Hough is used to transform correct the angle of plate. All steps are operated on MATLAB. Test result shows that, 94% of images which can not be segmented in unchanged algorithm is segmented successfully in algorithm changed.

关 键 词：字符分割 投影法 HOUGH变换 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]