摘要
基于开源视觉库Open CV,对行车环境中的路面交通标志的识别进行研究.在进行了适应于本次实验图片的预处理之后,先利用极角约束条件下的Hough变换检测出行车方向中的车道线,然后以车道线为基准进行路面交通标志在图片中位置的确定,进而提取交通标志的HOG特征,并结合SVM实现了路面交通标志的识别.实验结果表明,所提方法能较好识别行车中遇到的交通标志,具有一定的准确性、可靠性和鲁棒性,并且该方法不仅能处理静态图片,还能很好的应用于行车视频当中.
Based on the open computer vision library and preprocessing on the image recognition of road traffic signs in driving environment,the lane is detected according to the area of the polar angle constrains,the position of the road traffic signs in the picture is located. Later,the HOG features of the traffic signs are extracted and combined with SVM to recognize the class of the road traffic signs. Experimental results show that the proposed method can identify the encountered traffic signs with some accuracy,reliability and robustness,and the method can not only handle static image,but also be good for driving video.
出处
《大连交通大学学报》
CAS
2016年第3期103-106,111,共5页
Journal of Dalian Jiaotong University
基金
国家自然科学基金资助项目(51376028)
国家"863"计划资助项目(2013AA041108)
关键词
车道线检测
路面交通标志识别
导向标线
lane detection
road traffic sign identification
oriented marking
作者简介
刘萌雅（1989-），女，硕士研究生；
张丽艳（1972-），女，副教授，博士，主要从事信号处理与模式识别的研究E-mail：DALIANZHANGLIYAN@163．com
