Dodgebot

Summary:

Developed a face tracking system for a ball launcher robot, utilizing image data from ESP32-CAM and PID control to track smiling faces and throw balls at them.

Role and Responsibilities:

• Designed the face tracking algorithm using image data from the ESP32-CAM
• Implemented PID control to adjust the launcher’s position based on face detection

Skills and Tools Used:

• Programming: C++ (Arduino IDE), Python
• Libraries: RetinaFace (face detection), MQTT (data transmission)
• Hardware: ESP32-CAM, DC motor
• Fabrication: Laser-cut wooden base and launcher

Problem Statement:

The task was to create a robot capable of tracking smiling faces and launching balls at them, requiring accurate face detection and smooth motion control.

Solution Approach:

• Image Processing: Utilized the RetinaFace library to detect the center of a face, sending positional data from a Python script via an MQTT server
• Control: Implemented PID control in C++ to ensure precise motor adjustments, allowing the launcher to track and follow the target’s face
• Integration: Managed communication between the face detection system and the ESP32, ensuring seamless control over the ball-launching mechanism

Results and Impact:

• Achieved smooth, real-time face tracking and accurate ball launching
• Enhanced skills in PID control, image processing, and real-time data communication between devices

Reflection:

This project sharpened my abilities in robotics, particularly in integrating image processing with real-time control, and strengthened my experience in working with IoT protocols like MQTT