Eco-Vison

It's a prototype system that combines AI vision, OCR, and simulated IoT to make waste sorting smarter and more engaging. Five core pieces:

AI Waste Classification User uploads a photo (or uses a camera), and the system detects what type of waste it is, sorting into four categories: Biodegradable, Recyclable, Hazardous, or E-Waste. Powered by YOLOv8 object detection.
Hazard Detection via OCR For items with labels (chemical containers, batteries, etc.), EasyOCR reads the text on the label and scans for danger keywords — Toxic, Corrosive, Flammable, Chemical, Acid — and surfaces a warning if found.
Smart Bin Monitoring Simulates IoT fill-level sensors on bins, showing current fill percentage and predicting when a bin will overflow, so collection can be scheduled proactively instead of reactively.
Eco Rewards System Gamifies recycling — users earn Eco Points for correctly sorted waste, with a leaderboard and achievement badges to drive engagement.
Analytics Dashboard Aggregates the data: total waste processed, breakdown by category, hazardous-waste counts, and overall recycling rate.

Tech stack

Python + Flask backend, SQLite database, HTML/CSS/JS frontend (green eco theme), OpenCV + YOLOv8 for vision, EasyOCR for label reading