Calorie Camera

Mobile-first PWA that estimates calories from a food photo with editable results and uncertainty ranges.

Features

• Camera capture + upload fallback
• Two-stage analysis: vision detection + portion/nutrition estimate (plate cues or defaults)
• Editable foods and portion controls
• Local history stored in localStorage
• Friendly confidence banner + disclaimer

Tech

• Next.js App Router + TypeScript
• Tailwind CSS
• OpenAI Vision (optional) + USDA FoodData Central (optional)

Setup

npm install
npm run dev

Create a .env.local file:

OPENAI_API_KEY=your_openai_key
USDA_API_KEY=your_usda_key

If API keys are missing, the app falls back to manual entry and a local nutrition table.

Scripts

• npm run dev - local dev server
• npm run build - production build
• npm run start - run production build
• npm run test - unit tests for portion/calorie logic

Deploy

Vercel is recommended. Set OPENAI_API_KEY and USDA_API_KEY in the project environment variables.

Disclaimer

Estimates only, not medical advice.