🦷 Teeth Segmentation Mobile App (Flutter + YOLOv11 Backend)

This project implements a cross-platform mobile application for teeth segmentation using a YOLOv11 instance segmentation model. The solution is built in Flutter for the mobile front-end and uses a Flask-based backend for running the trained ML model. It accurately detects and outlines individual teeth from images and displays the segmentation in real time.

---

📹 Demo & Walkthrough

Watch the full project walkthrough on Loom:
👉 Loom Video Explanation

---

🔍 Project Structure

📱 Flutter Frontend

• Built using Flutter for both Android and iOS support.
• Captures or selects teeth images from gallery.
• Sends image to the backend API for segmentation.
• Receives polygon coordinates and renders them over the image using CustomPainter.
• Smooth UI and fast response thanks to server-side inference.

Key Features:

• Flutter UI/UX
• API integration with Flask server
• Real-time polygon overlay
• CustomPainter for drawing masks

🧠 ML Backend (Flask + YOLOv8)

• Built with Flask to expose a lightweight HTTP API.
• Loads a custom-trained YOLOv11 instance segmentation model.
• Accepts image input via POST request.
• Runs inference and converts model output to polygon coordinates.
• Returns JSON response with detected teeth regions.

Tech Stack:

• Flask
• YOLOv11 (Ultralytics)
• OpenCV, NumPy
• Polygon conversion utilities

---

🏋️‍♂️ Model Training Process

• We curated and annotated a dental image dataset using Roboflow.
• Our ML developer manually labeled segmentation masks for accurate training.
• Trained a YOLOv8 instance segmentation model optimized for tooth boundaries.
• Exported the final model weights for backend inference.

---

💡 Potential Use Cases

• Dental diagnosis assistance
• Smile design & cosmetic simulations
• Automated teeth labeling and counting
• Pre/post treatment comparison
• Dental record digitization

---

🚀 How to Run

Backend

1. Go to the Backend/ folder.
2. Install dependencies: Use PyCharm Editor
3. Run server: python app.py

Flutter App

1. Go to the Frontend/ folder.
2. Install packages: flutter pub get
3. Run app: flutter run

Make sure the backend is running and accessible by the mobile app.

---

📬 Contact

For any questions or collaboration inquiries, feel free to reach out via GitHub Issues or connect on LinkedIn.

• 🌐 Portfolio: offfahad.netlify.app
• 💼 LinkedIn: linkedin.com/in/offfahad
• 🐙 GitHub: github.com/offfahad