YOLOv11 Hailo Tracker

YOLOv11-Hailo-Tracker is a comprehensive real-time object detection, tracking, and speed estimation system optimized for Hailo AI accelerators. This project enables efficient detection of objects (with focus on persons and vehicles) with simultaneous tracking and speed calculation capabilities.

✨ Features

Real-time Object Detection: Using optimized YOLOv11 models for fast inference
Multi-Object Tracking: BYTE (ByteTrack) tracking algorithm for consistent object identification across frames
Speed Estimation: Advanced speed calculation for tracked objects using pixel-to-real-world distance conversion
Flexible Input Support: Images, videos, and camera streams
Configurable Labels: Support for detecting specific object classes (default: person, car)

🛠️ Requirements

Hardware

reComputer AI R2130

Purchase Now

reComputer AI Industrial R2135

Purchase Now

reComputer Industrial R2045-12

Purchase Now

reComputer Industrial R2045-12

Purchase Now

Camera or video input (for streaming applications)

Software

Python: 3.8 or higher
HailoRT: Version 4.23.0
Dependencies: Listed in requirements.txt

System Dependencies

Linux operating system (tested on Ubuntu)
OpenCV-compatible camera drivers (for camera input)
Graphics acceleration (for visualization)

📦 Installation

Prerequisites

Install Hailo PCIe driver and PyHailoRT from the Hailo website
Ensure your system has a compatible Hailo accelerator installed

Option 1: Local Installation

Clone the repository:

git clone https://github.com/Seeed-Projects/YOLOv11-Hailo-Tracker.git
cd YOLOv11-Hailo-Tracker

Create a virtual environment (recommended):

python -m venv venv --system-site-packages
source venv/bin/activate  # On Windows: venv\Scripts\activate

Install Python dependencies:
```
pip install -r requirements.txt
```
Run the project

  python run_api.py

Access localhost:8000 to reach the frontend and configure settings.

Option 2: Docker Deployment

Build and Run with Docker

# Build the Docker image with Hailo support
docker build --build-arg HAILO_VERSION=4.23.0 -t yolov11-hailo-tracker .

# Run the container
docker run -it --rm   -p 8000:8000   --privileged   --device /dev/hailo0:/dev/hailo0   -v /dev:/dev   yolov11-hailo-tracker

Using Pre-built Images

Pre-built Docker images with Hailo support are available as GitHub Releases and on GitHub Container Registry:

# Pull from GitHub Container Registry
docker pull ghcr.io/seeed-projects/yolov11-hailo-tracker-hailo:latest

# Run the container
docker run -it --rm   -p 8000:8000   --privileged   --device /dev/hailo0:/dev/hailo0   -v /dev:/dev   ghcr.io/seeed-projects/yolov11-hailo-tracker-hailo:latest

Build and Deploy with Docker Compose

Create a docker-compose.yml file:

services:
  yolov11-hailo-tracker:
    image: ghcr.io/seeed-projects/yolov11-hailo-tracker-hailo:latest
    container_name: yolov11-hailo-tracker
    privileged: true 
    ports:
      - "8000:8000"
    devices:
      - "/dev/hailo0:/dev/hailo0"
      - "/dev/video0:/dev/video0" 
    environment:
      - PYTHONUNBUFFERED=1
    command: python3 run_api.py
    restart: unless-stopped

Then run:

docker-compose up -d

🤝 Contributing

Fork the repository
Create a feature branch (git checkout -b feature/amazing-feature)
Commit your changes (git commit -m 'Add amazing feature')
Push to the branch (git push origin feature/amazing-feature)
Open a Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

YOLOv11 object detection models
BYTE (ByteTrack) multi-object tracking algorithm
Hailo Technologies for AI acceleration
OpenCV for computer vision operations

💞 Top contributors:

YOLOv11-Hailo-Tracker

Specs

README