Person Tracking Control of Mobile Robots Using a Lightweight Object Detection and Tracking System

This paper aims to develop a lightweight person tracking control system to facilitate real-time pedestrian tracking tasks for mobile robots. The proposed system utilizes deep learning technology and lightweight network architecture to perform pedestrian detection tasks to achieve powerful pedestrian detection capabilities. Subsequently, the integration of multi-object tracking technology enables the system to effectively track identified pedestrian targets, thereby enabling real-time pedestrian tracking of mobile robots. Experimental results show that the proposed lightweight person tracking control system exhibits robust pedestrian following performance, stability, and effectiveness under various scenarios and environmental conditions. These properties increase the potential of the proposed system in many practical applications.