Robotic target following with slow and delayed visual feedback
- PDF / 1,836,263 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 93 Downloads / 193 Views
REGULAR PAPER
Robotic target following with slow and delayed visual feedback Hui Xiao1 · Xu Chen1 Received: 15 July 2020 / Accepted: 14 October 2020 © Springer Nature Singapore Pte Ltd. 2020
Abstract Following rapidly and precisely a moving target has become the core functionality in robotic systems for transportation, manufacturing, and medical devices. Among existing targeting following methods, vision-based tracking continues to thrive as one of the most popular, and is the closest method to human perception. However, the low sampling rate and the time delays of visual outputs fundamentally hinder real-time applications. In this paper, we show the potential of significant performance gain in vision-based target following when partial knowledge of the target dynamics is available. Specifically, we propose a new framework with Kalman filters and multi-rate model-based prediction (1) to reconstruct fast-sampled 3D target position and velocity data, and (2) to compensate the time delay. Along the path, we study the impact of slow sampling and the delay duration, and we experimentally verify different algorithms with a robot manipulator equipped with an eye-inhand camera. The results show that our approach can achieve 95% error reduction rate compared with the commonly used visual servoing technique, when the target is moving with high speed and the visual measurements are slow and incapable of providing timely information. Keywords Visual servoing · Kalman filter · Delay compensation List of symbols {A} A 3D coordinate system with origin at porint A. A 𝜉B The 3D pose of frame {B} with respect to frame {A}. A vB The 3D velocity of frame {B} with respect to frame {A}. ⊕ The composition operator of relative poses, e.g., A 𝜉C = A 𝜉B ⊕B 𝜉C. A RB The rotation matrix corresponding to the relative pose A 𝜉B. A tB The translation vector corresponding to the relative pose A 𝜉B. Rx (𝜃) The rotation matrix that corresponds to the 3D rotating of 𝜃 degrees about x axis. A′ The transpose of matrix A.
* Xu Chen [email protected] Hui Xiao [email protected] 1
Mechanical Engineering, University of Washington, 3900 E Stevens Way NE, Seattle, WA 98195, USA
1 Introduction This paper considers the problem of controlling a robot to follow a moving target based on only visual feedback. Such capability relates to high-impact robotic applications ranging from autonomous ground or aerial vehicles that follow a leading target, to surgical robot arms that track the motions of human organs, and to robotic manipulators that perform pick-and-place tasks in a highly dynamic environment (e.g., above the sea or within a turbulence airflow). In those applications and the like, control based on visual feedback is a prime target for innovations. Indeed, vision sensors (e.g. cameras) are becoming ubiquitous and non-contact imaging is increasingly preferred in unstructured environments. In typical vision-based target following, image processing first extracts useful information from raw pixel data, then visual servo control algorithms calculate the
Data Loading...
