Research and Simulation of A Home Robot Navigation System Based on YOLO-SBA and ROS2

Le résumé fourni par la source

With the rapid advancement of smart homes and assistive care technologies, the application scope of domestic service robots in daily life has been continuously expanding.To ensure these robots can efficiently execute tasks and adapt to complex environments, their autonomous navigation capabilities have become a critical challenge.In home environments, robots must navigate dynamic obstacles, narrow passages, and variable environmental factors, which impose higher demands on their path planning and real-time decision-making abilities.To address these challenges, this study proposes an efficient navigation system for domestic service robots, integrating the YOLO-SBA object detection algorithm with the ROS2 robotic operating system.This research employs the ROS2 operating system for robot modeling and simulation, utilizing the COCO128 dataset for model training.Compared to the baseline YOLOv8n model, the YOLO-SBA model achieves a 2.1% increase in precision, a 10.2% improvement in recall, a 9.9% enhancement in mAP50, and a 9.3% boost in mAP50:95.The system's effectiveness was validated through laser SLAM mapping experiments conducted on the Gazebo simulation platform.The results demonstrate that the proposed system can efficiently perform object detection and SLAM mapping tasks in complex home environments.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.