A surveillance & security computer vision project.
Gunsense is licensed under ___; a non-restrictive, copy-left license.
This project CANNOT be relied upon for realworld surveillance situations. As it stands, this code is intended for hobby and research purposes.
- A CUDA-compatible GPU for effective processing capabilities
- Python >= 3.10
Gunsense can be installed in a few easy steps.
In the 21st century, gun violence is a major issue across the globe (but more specifically in the United States). While we can and should work towards preventing these horrible crimes, we should also focus on developing technology to effectively respond to them.
This project is pure, 100% Python.
I'm primarily using the three following frameworks/libraries/APIs:
- OpenCV -
- MediaPipe -
- Tensorflow -
There are three types of gunsense loggers:
- Stdout
- Error, event, and threat logging (in stdout/stream)
- File
- Error, event, and threat logging (on disk)
- Image
- Timestamped images of potential threats
- Stored alongside the text file log