Machine Vision Practice

Experiment 4 - YOLOv5 Object Detection

Experiment steps:

1. python3, opencv, and a conda environment must be installed. (Skip if already installed; refer to the environment installation flow in Experiments 1-3.)
2. Clone the YOLOv5 model. In the terminal, enter: git clone https://github.com/ultralytics/yolov5

3. cp -r /home/sunrise/yolov5 /home/sunrise/OPENCV/ # Copy the yolov5 package into the function package directory

cd OPENCV
pip install -r /home/sunrise/OPENCV/requirements_yolov5_torch.txt

(Installs the minimal dependencies for running YOLOv5)

4. source rdkx5_vision_env/bin/activate # Activate the virtual environment

5. Install the related packages:

In the OPENCV directory,

first upgrade pip: python -m pip install --upgrade pip

Run the following commands:

pip install torch torchvision --extra-index-url https://download.pytorch.org/whl/cpu）
pip install ultralyticspip install pandas psutil thop scipypython -m
pip install tqdm

Likely needed later; recommended to install together: python -m pip install pandas psutil thop pillow pyyaml requests matplotlib seaborn

(The following steps can be skipped initially. First try running the example file. If a too-low dependency version prevents the model from running, then update:

Upgrade the base install tools: python -m pip install -U pip wheel setuptools==70.0.0

Install/upgrade missing dependencies: python -m pip install -U gitpython pillow==10.3.0

)

1. cd yolov5 # Enter the package
2. python detect.py --weights yolov5s.pt --source 0 # Run the camera-based example; ensure the camera is properly connected

Terminal output:

If no camera is available, you can use a local image or video:

• Quick verification with a local image

• python detect.py --weights yolov5s.pt --source path\to\image.jpg # Replace path\to\image.jpg after --source with your image path

• Verification with a video file

• python detect.py --weights yolov5s.pt --source path\to\video.mp4 # Replace path\to\video.mp4 after --source with your video path