Unlicensed taxi detection algorithm based on traffic surveillance data.
Geographic Information Systems(2019)
摘要
ABSTRACTUnlicensed taxi services severely disrupt traffic management and threaten passengers' safety. Traditional approaches to detect unlicensed taxis rely highly on manual work like collecting on-site evidence. However, these approaches are ineffective and inefficient. As to address this hardship, we propose an unlicensed taxi detection algorithm using pass-records data collected from surveillance cameras. First, based on spatio-temporal analysis, we propose path irregularity and time irregularity to distinguish commercial vehicles from non-commercial vehicles. Then, we evaluate the algorithm based on the actual vehicle pass-records data of Guiyang. The results show that our algorithm outperforms baselines in terms of accuracy and running time.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要