Project overview
A geo-tagging system for mapping private cameras to aid law enforcement in crime investigations and enhance public safety with advanced video analytics.
Rajasthan Police Hackathon 1.0 put out twelve problem statements. Number 6 asked for a way to geo-tag privately owned cameras, so investigators could see which cameras cover an area, what each one can actually see, and who to contact for the footage.
I led the team that took it from selection through to the final stage at the grand finale in Jaipur, and the AI work was mine: assembling and cleaning the training dataset, then training the detection model that reads number plates and incidents such as fire and accidents out of camera footage. That was the bulk of my time on it.
The rest was filling gaps wherever the team needed them. I wired up the Twilio alerting so a match could reach an officer rather than sit in a dashboard, and worked across both the frontend and the backend as the two halves came together under a hackathon deadline.
Project gallery
Click any image to enlarge.