Blind Hat: Navigation Assistant for the Visually Impaired

🧭 Description

Blind Hat is a real-time smart navigation assistant that helps visually impaired individuals by guiding them safely using computer vision and GPS. It uses advanced obstacle detection and voice-based instructions to help users move confidently and independently.

🌐 Powered by a Flask server, the system runs locally or on GPU-enabled cloud platforms for real-time performance.

---

✨ Features

• Real-time Obstacle Detection with YOLOv8
• GPS-based Navigation with dynamic route updates
• Text-to-Speech Voice Alerts for instructions & warnings
• RESTful Flask API for integration with mobile or wearable clients
• GPU Acceleration Support for high-performance processing

---

🛠️ Technologies Used

Tech	Purpose
🐍 Python 3.8+	Backend & core logic
🔥 PyTorch	Deep learning model (YOLOv8)
📷 OpenCV	Video frame processing
🧠 YOLOv8	Object detection
🌐 Flask	Web server & APIs
🗺️ OSMnx	Route planning & shortest path
🗣️ pyttsx3	Text-to-speech
📡 Geocoder	Location via IP

---

🚀 Installation

1. Clone the Repository

   git clone https://github.com/iot-lab-kiit/BlindHat.git
   cd BlindHat-main

2. (Optional) Create a virtual environment

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

3. Install Dependencies

   pip install -r requirements.txt

4. Download YOLOv8 Model

   wget https://github.com/ultralytics/assets/releases/download/v0.0.0/yolov8s.pt

---

▶️ Usage

1. Start the Flask server:

   python app.py

2. Access the server at:

   http://0.0.0.0:5000
   

---

📡 API Endpoints

🎥 GET /video_feed

• Returns a real-time video stream (JPEG-encoded) with obstacle detection.

---

🔊 GET /audio_instructions

• Returns the latest voice instruction.

Response:

{
  "instruction": "Continue straight for 100 meters"
}

---

📍 POST /update_gps

• Updates the current location for route planning.

Request:

{
  "latitude": 20.348865,
  "longitude": 85.816085
}

Response:

{
  "status": "success"
}

---

☁️ Deployment

1. Choose a cloud provider (AWS, GCP, Azure)
2. Spin up a GPU-enabled VM (CUDA supported)
3. Install dependencies & upload the project
4. Run the Flask server:

   python app.py

5. Ensure secure access with HTTPS & authentication for production use

---

📱 Client-Side Integration

To use Blind Hat from a client device like a mobile app or wearable:

• Send captured video frames to the server
• Periodically update the GPS coordinates via /update_gps
• Fetch & play audio alerts from /audio_instructions
• Stream processed video from /video_feed

---

⚙️ Configuration

The destination coordinates are currently hardcoded in process_frame():

destination = (20.348865, 85.816085)  # KP 6

📍 Update this line to change the navigation destination.

---

📄 License

Licensed under the MIT License. See LICENSE.md for full details.

---

🙌 Acknowledgments

• YOLOv8 by Ultralytics
• OSMnx by Geoff Boeing
• Flask team for their minimalistic web framework
• OpenCV community for making vision accessible