Fake Product Identification Using QR Code-Based Authentication System

Abstract

Counterfeit consumer goods pose significant
economic and safety risks worldwide. This paper
proposes a robust framework for fake-product
identification using QR codes augmented with
cryptographic signatures and contextual tamperdetection to provide fast, reliable authentication at point
of sale. Each genuine item is assigned a secure, digitally
signed QR payload that contains a product identifier,
issuance metadata, and a short public-key signature. A
lightweight mobile scanning application verifies the
signature against a distributed verification service and
cross-checks product metadata against an immutable
audit log. To detect physical tampering and copied QR
images, the system complements cryptographic checks
with visual-analysis models that inspect QR placement,
surrounding packaging features, and printing artefacts
using convolutional neural networks trained on genuine
and counterfeit samples. We implement an end-to-end
prototype that integrates QR-generation tools, a
verification API, and a user-facing mobile client.
Experimental evaluation on a curated dataset of
packaged products demonstrates that combining
cryptographic verification with visual tamper-detection
significantly reduces false acceptances compared to
signature-only methods. The proposed approach is
designed for low-latency field use, scalable deployment
across supply chains, and can be adapted to sectors
ranging from pharmaceuticals to luxury goods. Future
work will explore privacy-preserving ledger options and
large-scale pilot deployments.