A Candidate Region Generation Algorithm Based on Shadow underneath Vehicle

摘要：

The candidate region generation algorithm is an important part for vehicle detection process. In order to further improve the reliability of the vehicle detection system, this paper presents a candidate region generation algorithm based on shadow underneath vehicle combined with edge feature and determined the region of interest. In addition, the algorithm can distinguish the shadow underneath vehicle and pavement domain by using adaptive threshold. Finally, the experiment results show that the novel algorithm can provide better recognition ratio and improve the reliability of the detection system.

关键词： vehicle candidate region region of interest (ROI) shadow underneath vehicle horizontal edge

作者: Juan Wang Xin Song Qiushi Zhang

作者单位: Computer Center, Northeastern University at Qinhuangdao, Qinhuangdao, Hebei, China

会议类型: 国际会议

会议名称: 2010 International Conference on Circuit and Signal Processing(2010年电路与信号处理国际会议 ICCSP 2010)

会议地点: 上海

会议语种:英文

页码: 107-110

在线出版日期: 2010-12-25（万方平台首次上网日期，不代表论文的发表时间）