基于Yolo V4的拥挤路段车辆检测

Vehicle detection in crowded Road Section based on Yolo V4

如今在车流量较为大的地方是需要对车辆进行检测监控,但这项任务艰巨,因此时使用智能化的方式去完成任务是较好的选择。YOLO V4算法是在YOLO V3的算法上添加许多ticks同时对backbone也进行了修改、增加了SPP模块增大模型的感受野。YoloV4在精确度和速度上都有所提升,不仅可以在拥挤路段对车辆完成静态识别,同时也可以完成实时检测。

Nowadays,in the place where the traffic flow is relatively large,it is necessary to detect and monitor the vehicles,but this task is difficult,so it is a better choice to use an intelligent way to complete the task.YOLO V4 adds ticks to THE algorithm of YOLO V3.Meanwhile,backbone is modified by adding SPP module to increase the sensing field of the model.YoloV4 has improved accuracy and speed,enabling static identification of vehicles on crowded roads as well as real-time detection.