YOLOv7-tracker 目标追踪 输入视频帧
·
参考项目:https://github.com/JackWoo0831/Yolov7-tracker/tree/master
github链接:https://github.com/Whiffe/Yolov7-tracker
码云链接:https://gitee.com/YFwinston/Yolov7-tracker
b站视频:https://www.bilibili.com/video/BV1Rh4y1K7zt/
1 项目安装
1.1 环境搭建
平台:AutoDL
选择镜像:
PyTorch 1.7.0 Python 3.8(ubuntu18.04) Cuda 11.0
1.2 项目下载
通过github下载
git clone https://github.com/Whiffe/Yolov7-tracker
或者通过码云下载(推荐)
git clone https://gitee.com/YFwinston/Yolov7-tracker
1.3 权重下载
下载权重,放在:Yolov7-tracker/weights中
权重下载:https://download.csdn.net/download/WhiffeYF/88250387
1.4 环境安装
pip install -r requirements.txt
pip install Cython
pip install cython_bbox
pip install motmetrics
1.5 上传待检测的视频帧
2 视频帧检测与追踪
2.1 检测与追踪
python tracker/track_demo.py --obj ./detect_frames/ --save_txt True
2.3 结果
结果有文本结果和图片结果:
文本结果如下:
Yolov7-tracker/demo_result/detect_frames/01
图片结果如下:
Yolov7-tracker/demo_result/result_images
更多推荐
已为社区贡献11条内容
所有评论(0)