Inspiration
Rodent infestations cause serious hygiene and safety problems, especially in food industries and warehouses. We wanted to create an AI-based solution that can detect rats in real-time using existing CCTV systems — with no extra hardware or software installation required.
What it does
Rat Detector is an AI-powered detection system that identifies rats directly through CCTV video feeds.
Detects and alerts when a rat is spotted in live footage.
Works seamlessly with existing CCTV infrastructure.
Requires only the source code — no additional setup or installation.
🧱 How we built it
Used Python and OpenCV for real-time video analysis.
Trained a deep learning model (YOLO-based) to recognize rats in various lighting and motion conditions.
Optimized the model for edge devices to ensure smooth CCTV integration.
Implemented a lightweight alert system to notify users instantly when a rat is detected.
🚧 Challenges we ran into
Training the model with limited and noisy rat image datasets.
Ensuring real-time performance without additional hardware.
Reducing false detections in dark or cluttered environments.
Integrating AI detection with live CCTV feeds efficiently.
🏆 Accomplishments that we're proud of
Built a fully functional detection model that runs on CCTV feeds with zero installation.
Achieved high accuracy and fast response times on basic hardware.
Designed a scalable system that can be implemented in factories, restaurants, or warehouses.
📚 What we learned
Handling real-time video data efficiently using computer vision.
Training and optimizing YOLO models for specific use cases.
Improving accuracy with limited datasets using augmentation.
Building practical AI systems that integrate directly with existing setups.
🚀 What's next for Rat Detector
Develop a web-based dashboard for live monitoring and reporting.
Log in or sign up for Devpost to join the conversation.