FreshScan AI 🥬📸

FreshScan is a modern, AI-powered Progressive Web App (PWA) built to analyze the freshness of raw food in real-time. By utilizing the device's camera and Google's Gemini Vision API, the app instantly identifies raw food (including specific meat cuts) and categorizes it as Fresh, Expiring Soon, or Spoiled.

🚀 Key Features

• Live Camera Integration: Native WebRTC camera capture within the browser.
• Gemini Vision AI: Zero-shot image evaluation powered by gemini-flash-lite-latest using dynamic markdown prompts.
• Edge Case Handling: Smart detection for NOT_FOOD (laptops, faces) and UNCLEAR (blurry/dark) images.
• Offline-First History: Guests save scans to localStorage. Once authenticated via Google, data automatically bulk-syncs to Firebase Firestore.
• Internationalization (i18n): Full support for English, Chinese Simplified (zh-CN), and Chinese Traditional (zh-TW).
• Progressive Web App (PWA): Installable on mobile devices with a standalone native feel.

🏗️ Architecture & Tech Stack

• Framework: Next.js 14 (App Router)
• Deployment: Vercel (Serverless API Routes)
• State Management: TanStack Query (React Query)
• Database & Auth: Firebase (Firestore + Authentication)
• Design: TailwindCSS v4 + Lucide React (Glassmorphism UI)
• Patterns: Object-Oriented Data Access (ScanRepository.ts), Absolute Imports (@/*), Context Providers (AuthContext.tsx).

🛠️ Getting Started

1. Environment Variables

Create a .env.local file in the root directory:

GEMINI_API_KEY=your_gemini_api_key_here

NEXT_PUBLIC_FIREBASE_API_KEY=your_firebase_api_key
NEXT_PUBLIC_FIREBASE_AUTH_DOMAIN=your_firebase_auth_domain
NEXT_PUBLIC_FIREBASE_PROJECT_ID=your_firebase_project_id
NEXT_PUBLIC_FIREBASE_STORAGE_BUCKET=your_firebase_storage_bucket
NEXT_PUBLIC_FIREBASE_MESSAGING_SENDER_ID=your_firebase_messaging_sender_id
NEXT_PUBLIC_FIREBASE_APP_ID=your_firebase_app_id

2. Run Locally

npm install
npm run dev

Open http://localhost:3000 with your browser to see the result.

3. Firestore Security Rules

To enable cloud syncing, ensure your Firestore database is created and deploy the following rules:

rules_version = '2';
service cloud.firestore {
  match /databases/{database}/documents {
    match /scan_history/{docId} {
      allow read, write: if request.auth != null && request.auth.uid == resource.data.userId;
      allow create: if request.auth != null;
    }
  }
}