基于速度分类算法的交通事件视频检测系统设计

摘要

Real-time video traffic incident detection method was proposed based on speed classification algorithm. In addition , traffic detection method, vehicles cross-road processing, speed detection, traffic flow detection and the identification of traffic events were also discussed. Based on vehicle detection and tracking, events such as traffic stop, lane transform times, slow traffic congestion and others can be identified and detected automatically to derive traffic flow, occupation ratio, queue length, average speed and other transportation parameters. In comparison with the traditional traffic incident detection system, the system is intuitive convenient and low-cost,and has good market demand and practical value.%提出基于速度分类算法的交通事件实时视频检测方法,并对交通量检测方法、车辆跨道处理、速度检测、交通状况检测及交通事件识别等进行了研究.在车辆检测与跟踪的基础上,可实现车辆停止、慢行、车道变换次数和车流拥挤等交通事件识别功能,通过自动检测车辆避障、车道变换、超速、慢速、停止和交通阻塞等事件,获得交通流量、占有率、排队长度、车型和平均车速等交通参数.与传统交通事件检测系统相比,具有直观方便、费用低等优点.