Indoor localization and tracking with RFID signal and video fusion for ROS-based mobile robot

Abstract

This study explores the potential of RFID (Radio Frequency Identification) technology for indoor localization and tracking, utilizing phase unwrapping and geometric localization techniques to achieve accurate 2D positioning. The research examines the localization accuracy achievable with RFID signals and integrates them with computer vision (CV) for enhanced target tracking. A tracking system combining CV and RFID technologies is designed and developed within the Robot Operating System (ROS) framework. By fusing RFID and CV data, the system enables unique and continuous target tracking in complex and cluttered environments, overcoming the limitations of CV, such as poor lighting, occlusion, and visually similar objects. Additionally, a two-layer obstacle avoidance strategy is implemented to ensure safe navigation in dynamic scenarios. Experimental results demonstrate the effectiveness of the proposed multi-sensor fusion approach, achieving robust and accurate tracking performance in varied indoor conditions. This work contributes to advancing multi-sensor fusion techniques for mobile robotics, addressing critical challenges in indoor localization, target tracking, and obstacle avoidance.