Skip to content
This repository has been archived by the owner on Aug 19, 2022. It is now read-only.

Extract vehicle trajectory/driver control information from a traffic video.

Notifications You must be signed in to change notification settings

neozhaoliang/visual-vehicle-behavier-analyzer

Repository files navigation

单目纯视觉分析车辆行为

这个项目演示通过用一部手机,拍摄道路上的一段交通视频,可以提取哪些有用的信息。

运行方法:

  1. 首先将 YOLOv3 的权重文件 yolov3.weights 下载并放在 yolov3_coco 目录下。
    wget https://pjreddie.com/media/files/yolov3.weights
    
  2. 运行 python main.py 查看效果。如果你的 Opencv 有 cuda 加速支持的话可以在 main.py 中将 use_gpu 设置为 True

详细文档见 doc 目录。

修复了 opencv 版本不同导致的轨迹不正确的错误。

About

Extract vehicle trajectory/driver control information from a traffic video.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages