基于伴生与互补颜色特征的车牌字符分割新方法

山东大学学报(工学版)

基于伴生与互补颜色特征的车牌字符分割新方法

王　枚¹^，²，王国宏¹

1. 海军航空工程学院电子通信系，山东烟台 264001；2. 烟台职业学院计算机及信息工程系,山东烟台 264000

收稿日期:2006-08-18 修回日期:1900-01-01 出版日期:2007-02-24 发布日期:2007-02-24
通讯作者: 王　枚

New method the vehical license plate character segmentation based on concomitant and complementary color features

WANG Mei¹^,², WANG Guo-hong¹

1. Department of Electronics & Communication, Naval Aeronautical Engineering Institute, Yantai 264001, China; 2. Department of Computer and Information Engineering, Yantai Vocational College, Yantai 264000, China

Received:2006-08-18 Revised:1900-01-01 Online:2007-02-24 Published:2007-02-24
Contact: WANG Mei

摘要/Abstract

摘要： 针对车牌区域的伴生与互补的颜色特征，提出一套字符分割的新算法.该算法利用定位到的车牌区域的原始颜色特征，直接进行字符及背景的提取，完成2值化过程，避免了图像处理过程中带来的误差，因此产生的噪声较少；最后利用垂直积分投影进行字符分割.对200幅从交通卡口获取的真实的彩色图像进行试验，算法对车牌破损，污垢、铆钉及字符粘连、过饱和等造成的影响均不敏感，其准确性，鲁棒性及实用性均较好.

关键词: 车牌字符分割, 伴生, 互补, 颜色特征

Abstract: A new method for character segmentation is given depending on the concomitant and complementary color features in the license plate area. The character color areas and background color areas are directly extracted from the original image of the license plate area. This was finished by binary process. By these methods, the errors arising during the image processing can be avoided, and lessen the noise made. Finally, the characters are segmented by the vertical projection method. The experiments on 200 automobile real RGB color images, which are taken under various conditions from traffic stations, show high accuracy. Robustness, practicability and lower sensitivity to the effects of plate disrepair, dirt, rivet and characters juncture can be obtained.

Key words: concomitant, complement, color feature , vehicle license plate character segmentation

中图分类号:

TP391.41

王　枚,王国宏 . 基于伴生与互补颜色特征的车牌字符分割新方法[J]. 山东大学学报(工学版), 2007, 37(1): 31-34 .

WANG Mei,WANG Guo-hong . New method the vehical license plate character segmentation based on concomitant and complementary color features[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2007, 37(1): 31-34 .

参考文献

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed

[1]	王飞,宋士瞻,曹永吉,谢红涛,张新华,张健,肖天,赵雅文. 基于连续小波变换的风光发电资源多尺度评估[J]. 山东大学学报(工学版), 2018, 48(5): 124-130.
[2]	刘建美, 马帅奇. 基于静、动态虚拟行动的个体交通出行模型构建[J]. 山东大学学报(工学版), 2014, 44(6): 55-62.
[3]	包书哲1,2,朱月澴1,王春立1*. 基于fMRI的图像底层特征关注研究[J]. 山东大学学报(工学版), 2014, 44(1): 24-28.
[4]	李景辉,杨立才*. 基于多传感器信息融合的人体姿态解算算法[J]. 山东大学学报(工学版), 2013, 43(5): 49-54.