Recognition of Notice Marking before Pedestrian Crossing
An approach to recognize the notice marking before pedestrian crossing with the shape of white rhombus is proposed.The target was extracted through color segmentation,and Hough transformation was utilized to detect the edge lines of the target.Harris corner detection algorithm was further used to acquire the two comers with the maximum change of target angle,by which to examine whether or not lines detected by Hough transformation are edge lines of the target and to eliminate the pseudo edge lines.At last,the intersectant points of the detected edge lines were utilized to judge whether the target is a rhombus.20 images including rhombus notice marking in the scenes and another 20 images having no rhombus were tested.Only two images were detected in errors and the detection rate reaches 90 percents.Non-rhombus images can be judged accurately in the scene that there exist no rhombus targets.The results show that the method is feasible and can be used in traffic signs recognition.
Notice marking before pedestrian crossing Rhombus detection Hough transformation Harris comer detection
Ning Zhang Tiejun HE Zhaohui Gao Hui Chen
ITS Research Center of Ministry of Education,Southeast University Nanjing,Jiangsu,210096,China
国际会议
大连
英文
1426-1429
2008-07-27（万方平台首次上网日期，不代表论文的发表时间）