Fast Vehicle Track Counting in Traffic Video

Ruoyan Qi*, Ying Liu, Zhongshuai Zhang, Xiaochun Yang, Guoren Wang, Yingshuo Jiang

*此作品的通讯作者

科研成果: 书/报告/会议事项章节会议稿件同行评审

1 引用 (Scopus)

摘要

In order to reduce road congestion, measures such as setting smart signal time schedules are needed. Therefore, it is a key technology to effectively and accurately count the traffic flow of vehicles at various intersections in the surveillance video. The method we propose uses Scaled-YOLOv4 as the vehicle detector, and then implements the vehicle tracking based on the DEEP SORT algorithm. To improve the accuracy and efficiency of the system, we propose a strategy of dynamic frame skipping based on density. We also propose to set key areas, combined with the driving direction, angle, etc., to judge and count the behavior of the vehicle. Experiments show that our method improved system efficiency while remaining high accuracy.

源语言英语
主期刊名Database Systems for Advanced Applications. DASFAA 2022 International Workshops - BDMS, BDQM, GDMA, IWBT, MAQTDS, and PMBD, Proceedings
编辑Uday Kiran Rage, Vikram Goyal, P. Krishna Reddy
出版商Springer Science and Business Media Deutschland GmbH
244-256
页数13
ISBN(印刷版)9783031112164
DOI
出版状态已出版 - 2022
活动International Workshops on BDMS, BDQM, GDMA, IWBT, MAQTDS, and PMBD 2022, held in conjunction with the 27th International Conference on Database Systems for Advanced Applications, DASFAA 2022 - Virtual, Online
期限: 11 4月 202214 4月 2022

出版系列

姓名Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
13248 LNCS
ISSN(印刷版)0302-9743
ISSN(电子版)1611-3349

会议

会议International Workshops on BDMS, BDQM, GDMA, IWBT, MAQTDS, and PMBD 2022, held in conjunction with the 27th International Conference on Database Systems for Advanced Applications, DASFAA 2022
Virtual, Online
时期11/04/2214/04/22

指纹

探究 'Fast Vehicle Track Counting in Traffic Video' 的科研主题。它们共同构成独一无二的指纹。

引用此