---

🥗 FoodCalci - Your Personal Food Tracker

FoodCalci is a lightweight web app that uses AI image recognition to detect foods from uploaded images and estimate their calorie values by integrating with the USDA FoodData Central API.

Try out FoodCalci live:

👉 https://staging.d282j2d0ipgah1.amplifyapp.com/

Upload a photo of your food — get instant calorie estimates in your browser.

---

🚀 Features

• 📷 Upload any food image
• 🧠 Detects food items using AWS Rekognition
• 🔍 Estimates real calories using USDA FoodData Central API
• 📊 Shows top detected food labels with confidence scores
• 💾 Remembers last uploaded result with localStorage
• 🎨 Clean UI, fully responsive
• ♿ Keyboard accessible & ARIA-friendly

---

📦 Tech Stack

Feature	Tech Used
Image Detection	AWS Rekognition (via browser SDK)
Calories API	USDA FoodData Central API
Auth	AWS Cognito Identity (anonymous)
UI	HTML, CSS, JavaScript
Storage	localStorage

---

🖼️ Demo

Upload an image (e.g., of a banana, burger, salad) and view calories instantly. Example output:

Name of food     Confidence
Banana           100.00
Fruit            100.00
Food             100.00
------------------------------
Estimated Calories: 89 Kcal

---

🛠️ Setup Instructions

1. Clone or download this repository
2. Place your own background image as background.jpg in the same directory
3. Open index.html in your browser or use amplify to host on AWS.

Note: You must have internet access to load AWS + USDA APIs from CDN.

---

🔐 API Keys Required

1. USDA API Key Sign up at https://api.data.gov/signup and get a free API key. Replace the placeholder in index.html:

   const apiKey = 'YOUR_API_KEY_HERE';

2. AWS Cognito Identity Pool ID Used for anonymous Rekognition access. Replace this with your own identity pool:

   IdentityPoolId: 'Replace with your pool ID'

---

⚠️ Limitations

• Calorie values are typically per 100g (as provided by USDA).
• Rekognition may return non-food labels
• No server backend — everything runs client-side (keys are visible in browser).
• USDA database may not return exact matches for all labels.

---

📈 Future Improvements

• Filter non-food labels automatically
• Add portion-size selector (e.g., 1 apple, 1 slice, etc.)
• Use a backend to hide API keys and cache frequent foods
• Add nutrition breakdown (protein, carbs, fats)

---

📄 License

This project is open-source and MIT licensed.

---