Autonomous Drone Tracking

Autonomous Facial Tracking Drone using YOLO8 Nano for face tracking model and Tello SDK for controls. Manually Implementing PID controls for smoother turns and 9 Concurrent threads running in the backend to handle all requests at the same time.

Demo | Link to UI | Link to Video

Technical Stack

• Backend: Python 3.8+ with DJI Tello SDK
• Frontend: Next.js 13+ web application
• Computer Vision: YOLO (Ultralytics)
• Control System: Custom PID controller with dead zone filtering
• Hardware: DJI Tello drone

Prerequisites

• Python 3.8 or higher
• Node.js 16.0 or higher
• DJI Tello Drone (fully charged)
• WiFi capability (to connect to drone network)

Installation

1. Clone the Repository

git clone https://github.com/yourusername/gazer.git
cd gazer

Running Gazer

Hardware Setup

1. Fully charge your DJI Tello drone
2. Power on the drone
3. Connect your computer to the drone's WiFi network (TELLO-XXXXXX)

Start the Application

Terminal 1 - Backend:

cd drone_backend
python3 main.py

Terminal 2 - Frontend:

cd next-app
npm run dev