AI-Driven Safety Decision Making (Food & Cosmetics)

Project Overview

Technology Stack: Python OpenCV TensorFlow/PyTorch OCR Web / Mobile UI

Problem Statement:
Manual inspection of labels, expiry, and safety compliance is error-prone and slow.

AI Component:
AI models classify products, detect anomalies, and flag unsafe items using image/video inputs.

Solution Summary:
OCR extracts product info, image/video analysis detects anomalies, and AI classifies products as safe or unsafe.

Impact:
Enhances public health, ensures global safety compliance; scalable for governments, retailers, and manufacturers.

Sample Dataset: Product Master

ProductID Name Category Brand ExpiryDate Ingredients SafetyRating
101 Natural Face Cream Cosmetics GlowFresh 2025-06-12 Aloe, Vitamin E Safe
102 Protein Energy Bar Food Nutrimax 2024-01-02 Whey, Cocoa Warning
103 Herbal Shampoo Cosmetics GreenLeaf 2025-12-10 Neem, Hibiscus Safe
104 Chocolate Drink Food CocoaPlus 2023-11-11 Cocoa, Sugar, Milk Expired
105 Baby Powder Cosmetics SoftCare 2026-03-21 Cornstarch, Lavender Safe

Sample Dataset: Image/Video Anomaly Detection

ImageID ProductID DetectedIssue Confidence AI Flag
IMG_001 102 Damaged Packaging 92% Unsafe
IMG_002 104 Expired Label Detected 99% Critical
IMG_003 101 No Issues 97% Safe
IMG_004 103 Foreign Object Detected 88% Unsafe
IMG_005 105 No Issues 99% Safe

Sample Dataset: OCR Extracted Information

OCR_ID ProductID ExtractedText FieldDetected AI Validation Result
OCR_01 102 EXP: 2024/01/02 Expiry Date Valid
OCR_02 104 EXP: 2023/11/11 Expiry Date Expired
OCR_03 101 Ingredients: Aloe, Vitamin E Ingredients Safe
OCR_04 103 Herbal Shampoo Neem Name/Category Matches
OCR_05 105 Batch No: 78245 Batch Number Valid