Scorpions Robotics #7672

FRC Rapid React Vision Processing

We use the Open Source Computer Vision Library OpenCV with Cascade Classifier features to detect and track the target.

Detect the host computer's OS, hostname and IP address.
Initialize NetworkTables client and stream server, then connect to the "vision" table.
Define the camera according to the OS and parameters.
Define the cascade classifier path. (The classifier we used is trained to detect FRC Power Ports, and can be downloaded from here)
while True:
- Read the image from the camera.
- Detect the target using the cascade_classifier.detectMultiScale() method.
- If the target is detected, draw a rectangle around it.
- Display the image.
- Send the target's position to the NetworkTables.
- Wait for the next frame.

Clone the repository. git clone https://github.com/Scorpions-Robotics/vision-2022.git
Install the required dependencies. pip install --upgrade -r requirements.txt
On Linux you have to install extra dependencies.
- cd misc/bash/
- chmod +x *
- ./install_os_dependencies.sh

Copy settings.ini.template and rename it to settings.ini. Then edit it to your needs.
Run the vision.py file to start the vision processing.

If you have any suggestions, or found any bugs, please open an issue or create a pull request.
Don't hesitate to open an issue if you have any questions about the code.