Master Comparator (Fake Product Detection) AI-powered web application designed to detect counterfeit products. It uses Computer Vision and digital scanning to verify the authenticity of a suspect item by comparing it against a verified reference product.
✨ How It Works
The system evaluates products in two distinct phases:
-
Phase 1: Visual Verification** Shape Similarity: Uses the ORB (Oriented FAST and Rotated BRIEF) algorithm via OpenCV to detect and match key geometric features between the real and suspect products. Color Similarity: Converts images to HSV format and compares their color histograms to ensure the exact shades and dyes match the original.
-
Phase 2: Digital Verification** Barcode / QR Scanning: Uses
zxing-cppto scan product tags or boxes, extracting hidden digital signatures or URLs to verify the supply chain.
Based on these combined scores, the AI generates a final verdict of MATCH CONFIRMED or MISMATCH DETECTED.
🛠️ Tech Stack
Frontend: HTML5, CSS3, Vanilla JavaScript (Fetch API)
Backend: Python, Flask
Computer Vision: OpenCV (opencv-python-headless), Numpy
Decoding: ZXing-CPP
Deployment: Render