Advanced computer vision and deep learning technology empowering consumers and shop owners to verify product authenticity instantly. One scan could save a life.
In 2024, South Africa faced a devastating food poisoning crisis. Counterfeit products infiltrated our communities, with tragic consequences that demand immediate action.
SpazaSafe combines advanced computer vision with deep learning to detect counterfeit products in seconds. Simply scan, capture, and know.
Point your camera at the product barcode to identify it
Take a clear photo of the product packaging
Deep learning compares structure, geometry, and color
Instant confidence score with clear guidance
ResNet18 deep learning analyzes packaging structure and design elements that counterfeiters cannot perfectly replicate.
ORB keypoint detection identifies spatial distortions and logo placement errors invisible to the human eye.
HSV histogram comparison detects subtle ink and printing deviations that reveal counterfeit production.
Optimized for low-bandwidth environments and older devices commonly used in township communities.
Get instant verification with confidence scores and actionable guidance before you buy.
Comprehensive South African product database verified against authentic reference images.
Deep learning framework for ResNet18 feature extraction
High-performance async Python API server
Computer vision for ORB features and HSV analysis
Cross-platform mobile app with OTA updates
SpazaSafe is made possible by the incredible work of open source communities and open data initiatives. We gratefully acknowledge their contributions.
Free, open, collaborative database of food products. Our product database is built on their incredible work.
openfoodfacts.orgOpen source machine learning framework powering our ResNet18 visual analysis engine.
pytorch.orgModern, fast web framework for building APIs with Python 3.7+ based on standard type hints.
fastapi.tiangolo.comOpen source computer vision library enabling our ORB feature detection and color analysis.
opencv.orgOpen source platform for building native apps using React, enabling rapid development and OTA updates.
expo.devEnabling the sharing and use of product images through CC BY-SA 3.0 licensing framework.
creativecommons.orgFounder of V3X
Driven by a passion for technology that creates real-world impact, Babongile developed SpazaSafe to address the urgent counterfeit food crisis affecting millions of South Africans. This project represents the intersection of advanced AI technology and community-focused solutions.
This project uses open data under the following licenses:
Product database from Open Food Facts is available under the Open Database License.
Product images are licensed under Creative Commons BY-SA 3.0 Unported.
Individual database contents available under Database Contents License.
Contains information from Open Food Facts, made available under the Open Database License (ODbL).