The purpose of this project is to create an analysis tool for any smart security system.
Current Features:
- Person detection
- Face recognition and grouping
Proposed Features
These features are currently being worked on
- Car recognition (Remembers your cars, alerts you when suspicious cars are on camera) ~ WIP
- License plate recognition (Extracts license plates from videos) ~ WIP
Step 1. To setup this rep, you will need the following installed:
pip3 install detectron2 opencv-python face_recognition numpy
Step 2: Clone my repo
Step 3: Make a new folder called video
in the root directory of this repo
Step 4: Copy/move footage from your smart cameras to the video
folder you just created.
Step 5: Run python3 object_extraction.py
. This will extract any important features used for analysis.
Step 6: Run python3 face_detection.py
. This will detect and label faces. Each detected face will be a folder within faces
(i.e. group0, group1, ..., etc.).
Step 7: More features to come!