摘要
限速标志的识别是智能交通系统的重要环节.模板匹配法在目前的交通标志识别领域中应用比较广泛,传统的模板匹配法对于限速标志的识别容易出现拒识和误识的问题,正确识别率不高.将改进模板匹配算法应用于限速标志的识别中,将限速标志字符的直观形象抽取特征,并结合边缘模板匹配,对限速标志进行识别,并在Vis-ual C++6.0环境下开发了限速牌识别软件系统.实验结果表明,基于改进模板匹配算法较传统模板匹配算法对限速标志的识别正确率有较大提高,识别率由80.95%提高到95.24%.
Identification of speed-limit signs is an important part of intelligent transportation systems. Template matching traffic sign identification is rather extensive in the current field of application. The traditional template matching identification of speed-limit sign is easy to error and to rejection. The correct identification rate is low. This will identify the speed-limit signs by means of the improved template matc- hing algorithm, and the speed limit signs will be taking the character of the visual image features, com- bined with the edge of the template matching, to identify the speed-limit signs, and to develop software system of identifying speed limit signs in Visual C++6.0 environment. Experimental results show that the improved template matching algorithm is better than the traditional template matching algorithm in identification of speed-limit signs, the identifying rate is improved from 80.95% to 95.24%.
出处
《西南大学学报(自然科学版)》
CAS
CSCD
北大核心
2013年第4期167-172,共6页
Journal of Southwest University(Natural Science Edition)
基金
国家大学生创新性实验计划(091063545)
关键词
限速标志
标志识别
模板匹配
字符特征
边缘提取
speed-limit sign
sign identification
template match
character feature
edge extracting