【目标跟踪】Yolov5_DeepSort_Pytorch复现

蔚落 2022-11-02 11:40 257阅读 0赞

# 1.环境 #

ubuntu16.04
    cuda10.1
    cudnn7
    python3.6
    
    
    Cython
    matplotlib>=3.2.2
    numpy>=1.18.5
    opencv-python>=4.1.2
    Pillow
    PyYAML>=5.3
    scipy>=1.4.1
    tensorboard>=2.2
    torch>=1.7.0
    torchvision>=0.8.1
    tqdm>=4.41.0
    seaborn>=0.11.0
    easydict

# 2.准备 #

## 2.0 准备代码与数据 ##

代码：

git clone --recurse-submodules https://github.com/mikel-brostrom/Yolov5_DeepSort_Pytorch.git

TownCentreXVID.avi测试视频数据：[https://download.csdn.net/download/qq\_35975447/12495835][https_download.csdn.net_download_qq_35975447_12495835]

## 2.1 重识别/分类模型 ##

在下面目录下载ckpt.t7

[https://drive.google.com/drive/folders/1xhG0kRH1EX5B9\_Iz8gQJb7UNnn\_riXi6][https_drive.google.com_drive_folders_1xhG0kRH1EX5B9_Iz8gQJb7UNnn_riXi6]

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70][]

放到目录下：

![20210225100648940.png][]

## 2.2 yolov5目标检测模型 ##

在下面目录下，下载一个最小的模型[yolov5s.pt][]。如果速度太慢，就右键复制链接，然后使用wget命令下载。

[https://github.com/ultralytics/yolov5/releases][https_github.com_ultralytics_yolov5_releases]

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 1][]

放到目录下：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 2][]

# 3.测试 #

## 3.1 修改 ##

将yolov5/utils/datasets.py中，所有的f"\{a\}"改为"\{\}".format(a)。

例如：

# line 1016
    assert cv2.imwrite(str(f), im[b[1]:b[3], b[0]:b[2]]), f'box failure in {f}'
    
    #改为：
    
    assert cv2.imwrite(str(f), im[b[1]:b[3], b[0]:b[2]]), 'box failure in {}'.format(f)

## 3.2 测试 ##

python track.py --source TownCentreXVID.avi  --save-txt

![20210225102851762.png][]

由于作者已经给出了动态gif，这里我只给个静态图。

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 3][]

[https_download.csdn.net_download_qq_35975447_12495835]: https://download.csdn.net/download/qq_35975447/12495835
[https_drive.google.com_drive_folders_1xhG0kRH1EX5B9_Iz8gQJb7UNnn_riXi6]: https://drive.google.com/drive/folders/1xhG0kRH1EX5B9_Iz8gQJb7UNnn_riXi6
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70]: /images/20221024/146e89f0407540f8a6ff6be35b56f7db.png
[20210225100648940.png]: /images/20221024/13f44f0afb9744a696a55d3577581a59.png
[yolov5s.pt]: https://github.com/ultralytics/yolov5/releases/download/v4.0/yolov5s.pt
[https_github.com_ultralytics_yolov5_releases]: https://github.com/ultralytics/yolov5/releases
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 1]: /images/20221024/858808e78f11491bba3c80ae8714c4f0.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 2]: /images/20221024/c402399f12a24d959a340bb3583ad05f.png
[20210225102851762.png]: /images/20221024/97b7a1e5c6ab4ca184c9eefea2e80295.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 3]: /images/20221024/df957e82c63a481991a9d143bd172320.png