Object Detection/Tracking Demo with YOLO and DeepSort - Soccer Ball Detection/Tracking

This demo uses the YOLO (You Only Look Once) model to detect soccer balls in images,videos and real-time webcam feeds. The project is implemented in Python and uses the ultralytics library for YOLO, along with OpenCV for image processing.

Project Structure

• env_yolo/: Virtual environment directory.

• output/: Directory for output files.

• README.md: This file.

• requirements.txt: List of dependencies.

• src/: Source code directory.

  • cam_test.py: Script to capture an image from the webcam.
  • soccer_ball.py: Script to detect soccer balls in images.
  • soccer_ball_cam.py: Script for real-time soccer ball detection using a webcam.
  • soccer_ball_video.py: Script to detect soccer balls in videos.
  • soccer_ball_track.py: Script to track soccer balls in videos using YOLO.
  • soccer_ball_deepsort.py: Script to track soccer balls in videos using DeepSort.

• yolo*.pt: YOLO model files.

Version Information

tested with

• CUDA Version: 12.6
• Python Version: 3.12.7

Setup

1. Clone the repository:

   git clone https://github.com/20wiz/yolo_basic
   cd yolo_basic

2. Create and activate a virtual environment:

   python -m venv env_yolo
   source env_yolo/Scripts/activate  # On Windows
   # source env_yolo/bin/activate    # On Unix or MacOS

3. Install the dependencies:

   pip install -r requirements.txt

   Depending on your GPU driver and CUDA version, you may need to reinstall PyTorch. https://pytorch.org/

4. YOLO model:

   yolo*.pt model file will be downloaded in the root directory of the project.

Usage

Detect Soccer Ball in an Image

1. Run the soccer_ball.py script:

   python src/soccer_ball.py

2. Modify the image_path variable in the script to point to your image file:

   image_path = '.\\test1.webp'  # Change to your image path

If you do not have a GPU available, you can set the device to CPU in the scripts by modifying the device variable:

device = 'cpu'  # Use CPU
# device = 'cuda:0'  # Use GPU if available

Real-Time Soccer Ball Detection with Webcam

1. Run the soccer_ball_cam.py script:

   python src/soccer_ball_cam.py

2. The script will start the webcam and display the real-time detection results. Press 'q' to quit.

Detect Soccer Ball in a Video

1. Run the soccer_ball_video.py script:

   python src/soccer_ball_video.py

2. Modify the video_path variable in the script to point to your video file:

   video_path = 'soccer_1.mp4'  # Change to your video path

Track Soccer Ball in a Video using YOLO

1. Run the soccer_ball_track.py script:

   python src/soccer_ball_track.py

2. Modify the video_path variable in the script to point to your video file:

   video_path = 'soccer_1.mp4'  # Change to your video path

Track Soccer Ball in a Video using DeepSort

1. Run the soccer_ball_deepsort.py script:

   python src/soccer_ball_deepsort.py

2. Modify the video_path variable in the script to point to your video file:

   video_path = 'soccer_1.mp4'  # Change to your video path

License

This project is licensed under the MIT License.

Acknowledgements

• Ultralytics YOLO
• OpenCV
• Claude