Replanejamento de Movimento Baseado em Visão da Tarefa Realizada por um Manipulador Robótico em uma Célula de Inspeção de Peças Automotivas

Abstract

Machine vision is one of the most commonly employed techniques for quality control in industrial processes, and the combination of robotic manipulators and machine vision systems grants increased flexibility and effectiveness of such systems. This paper presents a movement replanning strategy of a robotic manipulator to perform the visual inspection of an automotive part, based on information extracted from images. A robotic manipulator is responsible for positioning a camera in various poses in order to capture images of the inspected part. However, this system is prone to failure in case the part is not in the expected position. The proposed system’s objective is to avoid this imposition, by replanning the inspection poses based on the estimated part position. Experimental results show that the system was capable of performing successful inspections of the part, despite large pose variations of the part inside the workspace.