This project does ANPR and facial recognition using ultralytics YOLOv5 Object detection model. Inferencing is done on both images and video input.
This can be used to design an end-to-end system that detects faces and number plates from real time traffic feed. The detected faces, number plates enclosed within bounding boxes can be stored in a NOSQL database along with a timestamp and periodically sent to the law enforcement authorities.
-
Open the given ipynb file in google colab
-
Run it!
NOTE : we use detect-1.py which is a modified version of detect.py which crops out the images within bounding boxes.