Sweet - 2 DOF Object Tracking Pan and Tilt Mechanism

This project demonstrates a 2 Degrees of Freedom (2-DOF) pan and tilt mechanism designed to track and follow objects, such as faces, using deep learning. The system utilizes an Arduino for servo control, a main PC for object detection and localization, and a RealSense camera for capturing video streams.

View

Table of Contents

• Introduction
• Features
• Hardware Components
• Software Components
• Installation and Setup
• Usage
• Video Demonstration
• License
• Acknowledgements

Introduction

The objective of this project is to create a pan and tilt mechanism that can follow an object, specifically a human face. This is achieved by using two servo motors to control the horizontal (pan) and vertical (tilt) movement of the camera, an Arduino for servo control, and a main PC running a deep learning algorithm to detect and localize objects in real-time.

Features

• Real-time Object Detection: Utilizes OpenCV and Haar cascades for face detection.
• 2-DOF Control: Pan and tilt mechanism controlled by two servo motors.
• Arduino Integration: Arduino handles servo control based on commands from the main PC.
• 3D Printed Parts: Custom designed and 3D printed parts for mounting the camera and servos.

Hardware Components

• 2 x Servo Motors
• 1 x Arduino (e.g., Arduino Uno)
• 1 x Intel RealSense Camera
• 3D Printed Parts for mounting and mechanism
• Connecting wires and power supply

Software Components

• Python
• OpenCV
• pyrealsense2
• Arduino IDE

Installation and Setup

Hardware Setup

1. 3D Print Parts: Print all necessary parts for the pan and tilt mechanism.
2. Assemble Mechanism: Attach the servo motors and RealSense camera to the 3D printed parts.
3. Connect Servos to Arduino:
   • Connect the yaw (pan) servo to pin 9 on the Arduino.
   • Connect the pitch (tilt) servo to pin 10 on the Arduino.
4. Connect Arduino to PC: Connect the Arduino to your main PC via a USB cable.

Software Setup

1. Clone Repository:

   git clone [email protected]:abhismirai10/Sweet_a_2_dof_robotic_follower.git
   cd 2dof-object-tracking

2. Install Dependencies:

   pip install opencv-python pyrealsense2 numpy pyserial

3. Upload Arduino Code:
   • Open the Arduino IDE.
   • Load the servo_control.ino file from the arduino directory.
   • Select the appropriate board and port, then upload the code to the Arduino.

Usage

1. Run the Python Script:

   python object_tracking.py

2. Start the RealSense Camera: Ensure the camera is properly connected and functioning.
3. Face Detection: The system will start detecting faces and send servo control commands to the Arduino to adjust the pan and tilt angles accordingly.
4. Stop the Program: Press q in the OpenCV window to terminate the program.

Video Demonstration

Watch the video demonstration of the project on YouTube: 2-DOF Object Tracking Pan and Tilt Mechanism

License

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

Acknowledgements

• OpenCV for the computer vision library.
• Intel for the RealSense camera.
• Arduino for the microcontroller platform.
• Special thanks to the contributors of the project.

---

Feel free to contribute to this project by opening issues or submitting pull requests. For any inquiries or feedback, please contact [[email protected]].