PixaShield

PixaShield is an innovative AI-powered camera solution designed to enhance public safety and security surveillance. Developed for the Rajasthan Police Hackathon 1.0, this project earned first place in Problem Statement #3: "AI for intelligent camera decision making," winning a prize of ₹50,000.

The system leverages Ultralytics YOLOv8, trained on over 125,000 images across multiple datasets. Our model can detect various objects including people, vehicles, drones, and weapons like handguns, rifles, and grenades. What sets PixaShield apart is its ability to process video feeds with low latency while maintaining high accuracy in threat detection.

The Rajasthan Police Hackathon, held at the Rajasthan International Centre in Jaipur from January 17-18, 2024, was an intensive 36-hour event where our team worked tirelessly, often with just a few hours of sleep. The hackathon featured over 300 participating teams across 12 problem statements and included special events like drone shows and live performances.

As part of the development team, I focused on implementing the core computer vision algorithms and optimizing the machine learning models for real-time processing. Our solution integrated notification systems including email alerts and Twilio API for automated calls when potential threats are detected.

Demo video showcasing PixaShield's object detection capabilities