Vision-Based Control of a Mobile Robot: A Real-Time Approach

Abstract

This thesis presents the design and implementation of an autonomous mobile robot that uses computer vision to detect and follow a colored object (e.g. green ball) in real-time. The system employs an ESP32-CAM module for video streaming and an ESP32 microcontroller for motion control, receiving commands over Wi-Fi from a host computer running the object tracking algorithm. A GPS module was initially planned for absolute localization but was removed due to technical failures. The project demonstrates that with low-cost components and basic vision techniques, effective autonomous behavior can be achieved in a semi-structured environment.