Robust Vision-Based Pose Correction for a Robotic Manipulator Using Active Markers

Abstract

Robots with elastic or lightweight components are becoming common in research, but can suffer from undesired positioning imprecision, which motivates a vision-based pose correction of the manipulator. For robotic manipulators that operate outdoors and under changing illumination conditions, robustness of the vision components is of principal concern. We propose a monocular manipulator pose correction based on active markers which are detected by convergence criteria on the image gradient field. We show the capabilities of the method in several outdoor and indoor experiments, considering the use case of a planetary exploration rover prototype equipped with a lightweight robotic arm. The vision-based manipulator pose correction method proves to be successful despite back light, reflections, and image overexposure and additionally allows continued robot operation in the case of extrinsic camera decalibration.