In-Hand Object Pose Estimation via Visual-Tactile Fusion

TL;DR

This paper introduces a method that fuses visual and tactile data to improve in-hand object pose estimation for robotic manipulation, especially under occlusion, achieving higher accuracy than vision-only methods.

Contribution

It presents a novel sensor fusion approach combining RGB-D and tactile sensors with an adapted ICP algorithm for improved pose estimation accuracy.

Findings

Tactile data significantly improves pose accuracy under occlusion.

Achieved average pose error of 7.5 mm and 16.7 degrees.

Outperformed vision-only baselines by up to 20%.

Abstract

Accurate in-hand pose estimation is crucial for robotic object manipulation, but visual occlusion remains a major challenge for vision-based approaches. This paper presents an approach to robotic in-hand object pose estimation, combining visual and tactile information to accurately determine the position and orientation of objects grasped by a robotic hand. We address the challenge of visual occlusion by fusing visual information from a wrist-mounted RGB-D camera with tactile information from vision-based tactile sensors mounted on the fingertips of a robotic gripper. Our approach employs a weighting and sensor fusion module to combine point clouds from heterogeneous sensor types and control each modality's contribution to the pose estimation process. We use an augmented Iterative Closest Point (ICP) algorithm adapted for weighted point clouds to estimate the 6D object pose. Our experiments show that incorporating tactile information significantly improves pose estimation accuracy, particularly when occlusion is high. Our method achieves an average pose estimation error of 7.5 mm and 16.7 degrees, outperforming vision-only baselines by up to 20%. We also demonstrate the ability of our method to perform precise object manipulation in a real-world insertion task.