NutriVision AI

A web-based nutrition analyzer that uses Google Gemini to identify food from photos and track daily nutrient intake. Built with Flask and SQLite, designed for both casual users and those with specific fitness goals.

---

What It Does

Upload a photo of your meal — the app identifies the food, estimates portions, and returns a full nutritional breakdown covering calories, macros, vitamins, minerals, and water content. Everything gets logged to your daily tracker automatically.

There's also a built-in chatbot (NutriBot) that can answer nutrition questions, suggest meals, and comment on your daily progress based on what you've actually eaten.

Guests can try one free analysis without registering. Registered users get full history, personalized calorie targets, and the feedback/correction system.

---

Screenshots

Landing Page

Analysis Results

Nutrition Summary

Recommendations

NutriBot Chatbot

---

Tech Stack

• Backend: Python, Flask 2.3.3
• AI: Google Gemini 2.5 Flash
• Database: SQLite
• Auth: bcrypt + Flask session (7-day persistent)
• Frontend: HTML, CSS, Vanilla JS
• Other: Flask-Limiter, Pillow, Flask-CORS

---

Project Structure

nutrivision-ai/
├── main.py               # App entry point, all routes
├── auth.py               # Login, register, guest session handling
├── database.py           # DB setup and queries
├── nutribot.py           # Chatbot logic
├── config.py             # Config, nutrition database, daily targets
├── requirements.txt
│
├── templates/
│   ├── landing.html
│   ├── index.html        # Login/register
│   └── app.html          # Main dashboard
│
├── static/
│   ├── style.css
│   ├── landing.css
│   ├── script.js
│   ├── landing.js
│   └── nutrition-fix.js
│
└── docs/
    └── screenshots/
        ├── landing.jpg
        ├── analysis.jpg
        ├── summary.jpg
        ├── recommendation.jpg
        └── chatbot.jpg

---

Setup

1. Clone and enter the directory

git clone https://github.com/yourusername/nutrivision-ai.git
cd nutrivision-ai

2. Create a virtual environment

python -m venv venv
source venv/bin/activate        # Linux/macOS
venv\Scripts\activate           # Windows

3. Install dependencies

pip install -r requirements.txt

4. Create a .env file

SECRET_KEY=your-secret-key
GEMINI_API_KEY=your-gemini-api-key
DATABASE_PATH=nutrition_analyzer.db
GUEST_ANALYSIS_LIMIT=1
FLASK_DEBUG=True
SESSION_COOKIE_SECURE=False

Get your Gemini API key from Google AI Studio.

5. Run

python main.py

Open http://localhost:5000.

---

API Reference

Method	Endpoint	Auth Required	Description
POST	/api/auth/register	No	Register new account
POST	/api/auth/login	No	Login
POST	/api/auth/logout	No	Logout
POST	/api/analyze	Auth or Guest	Analyze food photo
GET	/api/history	Yes	Meal history
GET	/api/dashboard	Yes	Daily nutrition summary
POST	/api/update-daily-nutrition	Yes	Update daily totals
POST	/api/nutribot/chat	Auth or Guest	Chat with NutriBot
GET	/api/profile	Yes	Get user profile
PUT	/api/profile/update	Yes	Update profile
POST	/api/feedback	Yes	Submit feedback
POST	/api/feedback/correction	Yes	Correct food identification
POST	/api/feedback/rating	Yes	Rate an analysis
GET	/api/feedback/stats	No	Feedback statistics
GET	/api/health	No	Health check

Rate limits: 500/day · 100/hour · 20/minute per IP.

---

Default Nutrition Targets

These apply to guests and new users. Registered users get targets calculated from their profile (age, weight, height, activity level, fitness goal).

Nutrient	Default
Calories	2000 kcal
Protein	50 g
Carbohydrates	250 g
Fat	65 g
Fiber	25 g
Vitamin A	900 mcg
Vitamin C	90 mg
Calcium	1000 mg
Iron	18 mg
Water	2000 ml

Fitness goal adjustments: −500 kcal (lose weight), +300 kcal (build muscle), +500 kcal (gain weight).

---

Notes

• Max upload size: 16 MB
• Gemini API calls include automatic retry (up to 3 attempts) with 1-second delay
• Session cookies are HTTPOnly and SameSite=Lax by default; set SESSION_COOKIE_SECURE=True in production
• The built-in NUTRITION_DB in config.py covers common foods including Taiwanese street food, local vegetables, and seafood — Gemini handles anything outside the database
• Never commit your .env file — add it to .gitignore before pushing

---

Requirements

Python 3.9+

Flask==2.3.3
Flask-CORS==4.0.0
Flask-Limiter==3.5.0
bcrypt==4.0.1
PyJWT==2.8.0
google-generativeai==0.3.1
Pillow>=9.5.0,<11.0.0
python-dotenv==1.0.0
Werkzeug==2.3.7
requests==2.31.0

---

License

MIT