Finds unique objects in images and videos by matching a given image of the object. It uses OpenCV Feature Matching algorithms.

Can be used as an extra validation method for autonomous drone flights.

This project uses some licensed mathematical algorithms as an example of implementation. This is a research project and I am not responsible for anything.

Tested with Python v11, v13, v14. Use venv and pip-tools for correct versions matching. Create requirements.txt for your environment using pip-compile and pip-sync commands.

Check your PYTHONPATH and modify the imports if necessary.

data directory contains cases and configurations. Easiest is to copy one of the existing directories and update config.py and .json files.

Update path and arguments in main.py and video_main.py.

Run with:

python src/main.py  # or
python src/video_main.py

• Image enhancements
• SLAM
• Thermal images