基于字符块提取的车牌字符分割算法  

Character Segmentation Algorithm of Vehicle License Plate Based on Character Extraction n

作　　者：薛倩[1] 

机构地区：[1]陕西交通职业技术学院信息工程系,西安710018

出　　处：《河南科学》2014年第5期781-784,共4页Henan Science

基　　金：国家自然科学青年基金项目资助(41102107)

摘　　要：为了解决车牌图像倾斜、背景复杂、分割过程中出现的字符间粘连、断裂等问题,提出简便有效抗干扰强的基于字符块提取的车牌字符分割算法,以此提升车牌字符的识别效果.通过车牌图像二值化处理、倾斜矫正、去除干扰以及字符块提取一系列步骤,实现车牌识别前对车牌字符的准确有效分割.实验结果表明,该车牌字符分割方法可靠、准确度高,为后续车牌字符的正确识别奠定基础.In order to solve the problem of license plate image tilt, complex background, character adhesion or frac- ture in the segmentation process, a simple, effective and strong anti-jamming algorithm for segmentation of license plate characters based on character extraction is presented, so as to enhance the license plate character recognition effect. After multistep processes of binary processing license plate image, tilt correcting image, removing interfer- ence from image and character extraction, license plate characters are segmented accurately and effectively before the vehicle license plate recognition. The experimental results show that the license plate character segmentation method is reliable, accurate, and lays a foundation for the subsequent license plate character recognition.

关 键 词：车牌字符分割 字符块提取 二值化 HOUGH变换 去除干扰 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]