Hand-Eye Calibration of EOD Robot by Solving the AXB = YCZD Problem

Jianfeng Jiang, Xiao Luo*, Shijie Xu, Qingsheng Luo, Minghao Li

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)

Abstract

The explosive ordnance disposal (EOD) robots work in a special environment, which requires dual robotic arms to work together for removing the bomb. Therefore, machine vision is very important to locating bombs, especially the accuracy of the hand-eye calibration technology. A basic problem encountered with the collaborative work of dual robotic arms is to solve the unknown homogeneous transformation matrix, including: hand-eye of robotic arm 1, base-base, and camera-end effector's robotic arm 2. In this article, the hand-eye calibration problem of the dual robotic arm system is expressed as two matrix equations. A new method of simultaneously solving the unknowns in the matrix equation is proposed. This method consists of a closed-form method based on the Kronecker product and an iterative method that transforms a nonlinear problem with a convex function optimization problem. The closed-form method is used to quickly obtain the initial value of the iteration method to improve the efficiency and accuracy of the iteration. In addition, we propose a hand-eye calibration method based on the re-projection error of the RGB-D camera. In order to prove the feasibility and superiority in the proposed iterative method, we conducted simulation and actual experiments and compared them with the other two calibration methods. The comparison results verify the superiority in the proposed method in terms of accuracy.

Original languageEnglish
Pages (from-to)3415-3429
Number of pages15
JournalIEEE Access
Volume10
DOIs
Publication statusPublished - 2022

Keywords

  • EOD robot
  • Nonlinear optimization
  • RGB-D camera
  • Re-projection error
  • Simultaneous hand-eye calibration of the dual robot

Fingerprint

Dive into the research topics of 'Hand-Eye Calibration of EOD Robot by Solving the AXB = YCZD Problem'. Together they form a unique fingerprint.

Cite this