ROS-Enabled Automated Fruit Sorting with YOLOv5 Object Detection and Navigation Integration on a Wheeled Mobile Robot

Abstract

This paper introduces a ROS (Robot Operating System) based wheeled mobile robot system designed for automated fruit sorting utilizing YOLOv5 (You Only Look Once version 5) object detection, map building, and navigation functionalities. The integration of robotic systems with advanced computer vision techniques is an expanding research field that aims to improve efficiency, productivity, and sustainability in food production. The integration of these technologies enables the robot to autonomously detect, classify, and sort fruits in real-time while navigating within its environment. The YOLOv5 object detection model provides high-accuracy fruit detection capabilities, essential for precise sorting in dynamic and cluttered environments. The novelty of the work is in utilizing ROS for map building and navigation, leveraging SLAM (Simultaneous Localization and Mapping) techniques to create and update maps of its surroundings, enabling adaptive and efficient navigation. Our experimental results demonstrate the effectiveness of the proposed system in achieving accurate fruit sorting performance while navigating autonomously using hardware. The developed robotic system exhibits considerable potential for enhancing efficiency and productivity in agricultural and industrial fruit sorting operations, paving the way for future advancements in robotic automation.