https://tile-extractor.onrender.com/

🪨 Tile Image Extractor

A web application to automatically extract all tile product images from PDF catalogues — losslessly, at high speed, unlimited times.

Built by Somya Bhalani

---

✨ Features

• ⚡ Lightning Fast — Processes large PDF catalogues in seconds using async background tasks
• 🔍 Smart Tile Detection — AI-powered filter rejects QR codes, logos, borders, and blank fills — keeping only real tile images
• 🖼️ Lossless Quality — All images extracted without any quality loss (JPEG byte-copied, others converted to web-safe PNG)
• 📦 Download All — Export every extracted tile as a single ZIP archive in one click
• 📊 Live Progress Bar — Real-time extraction progress tracked from backend to frontend
• 🪟 Masonry Gallery — Beautiful responsive grid layout with fullscreen lightbox viewer
• 📱 Mobile Friendly — Fully responsive design, works on all screen sizes
• 🎨 Liquid Glass UI — Premium frosted glass design with marble background

---

🚀 Getting Started

1. Prerequisites

You will need the following installed on your machine:

• Python 3.9+ and pip

2. Installation

Open your terminal or command prompt and run:

# Clone the repository
git clone https://github.com/somyabhalani/tile-extractor.git

# Navigate into the folder
cd tile-extractor

# Install the required Python libraries
pip install -r requirements.txt

3. Run Locally

Start the local server using Uvicorn:

uvicorn app:app --reload

4. Open in Browser

Once the server is running, open your web browser and go to: http://localhost:8000

---

🛠️ Tech Stack

Layer	Technology
Backend	Python, FastAPI, Uvicorn
PDF Engine	PyMuPDF (fitz)
Images	Pillow (PIL)
Frontend	HTML5, CSS3, Vanilla JS
Design	Glassmorphism / Liquid Glass

---

📁 Project Structure

tile-extractor/
├── app.py               # FastAPI server & API endpoints
├── tile_extractor.py    # Core PDF tile extraction engine
├── requirements.txt     # Python dependencies
├── static/
│   ├── index.html       # Main UI
│   ├── style.css        # Liquid Glass design system
│   ├── script.js        # Frontend logic & progress polling
│   └── marble-bg.png    # Background texture
└── jobs/                # Auto-generated; stores extracted tiles per session

📖 How It Works

1. Upload — Drop a PDF tile catalogue onto the upload area
2. Extract — The backend spawns a background task, iterating through every image reference in the PDF
3. Filter — Each image passes through a smart TileFilter that rejects non-tile content based on size, aspect ratio, edge density, and color statistics
4. Progress — The frontend polls /api/progress/{job_id} every 500ms and animates a live progress bar
5. Display — Extracted tiles are served via /api/images and displayed in a beautiful masonry gallery
6. Download — Optionally download all tiles as a ZIP archive via /api/download/{job_id}

---

📞 Contact

Somya Bhalani

• GitHub: @somyabhalani
• Phone: +91 8320408204

---

© 2025 Somya Bhalani. All rights reserved.