摘 要:高速公路上经常出现车辆超速、违章停车、逆行、抛洒物等异常事件,为了在出现异常事件时能够自动快速识别并预警,通过硬件的物理部署、软件的搭建以及 YOLOv5 目标检测及 DeepSort 目标跟踪技术的研究及应用,开发出应用于高速公路异常事件的智能化自动化检测及预警系统。并通过落地案例验证方案的可行性及实用性,为高速公路异常事件的快速识别及预警提供借鉴和参考。
关键词:AI 视频识别;异常事件;YOLOv5 目标检测;DeepSort 目标跟踪;自动预警
DOI:10.19850/j.cnki.2096-4706.2022.24.005
中图分类号:TP18;U491.2 文献标识码:A 文章编号:2096-4706(2022)24-0018-05
Research on Highway Anomaly Event Monitoring System Based on AI Video Recognition Technology
CHEN Guohua
(Guangdong Feida Traffic Engineering Co., Ltd., Guangzhou 510663, China)
Abstract: The vehicle speeding, illegal parking, retrograde, throwing objects and other abnormal events often appear on the highway, in order to automatically quickly identify and conduct early warning when the abnormal events occur, through the physical deployment of hardware, software construction, YOLOv5 target detection, research and application of DeepSort target tracking technology, this paper develops an intelligent automatic detection and early warning system which is applied to abnormal events on the highway. It verifies the feasibility and practicality of the scheme through the landing cases, provides reference for the rapid identification and early warning of abnormal events on the highway.
Keywords: AI video recognition; anomaly event; YOLOv5 object detection; DeepSort object tracking; automatic early warning
参考文献:
[1] 魏溪含,涂铭,张修鹏 . 深度学习与图像识别原理与实践[M]. 北京:机械工业出版社,2019.
[2] 赵文杰,刘树潜,王菲,等 . 基于 AI 视频检测的高速公路事件分析探究 [J]. 中国交通信息化,2022(2):112-115+130.
[3] 李海生,蒋周龙,李杰 . 高速公路交通视频事件检测系统应用探讨 [J]. 中国交通信息化,2022(3):102-106.
[4] 杨莹 . 基于视频大数据的高速公路异常事件智能监测系统[J]. 电子世界,2020(11):99-100+103.
[5] 刘伟 . 高速公路视频事件检测系统研究 [J]. 智能城市,2021,7(10):123-124.
[6] 杨常华 . 基于深度学习的高速公路视频事件检测系统分析[J]. 山西电子技术,2020(2):13-14+86.
[7] 扶松柏.图像识别技术与实战 [M].北京:清华大学出版社,2022.
作者简介:陈国华(1972.02—),男,汉族,广东惠州人,中级工程师,本科,研究方向:智能交通。