Citation: Zhou, W.; Chen, H.; Jin, Z.;
Zuo, Q.; Xu, Y.; He, K. A Novel and
Simplified Extrinsic Calibration of 2D
Laser Rangefinder and Depth
Camera. Machines 2022, 10, 646.
https://doi.org/10.3390/
machines10080646
Academic Editors:
Antonis Gasteratos
and Luis Payá
Received: 14 June 2022
Accepted: 30 July 2022
Published: 3 August 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
A Novel and Simplified Extrinsic Calibration of 2D Laser
Rangefinder and Depth Camera
Wei Zhou
1,2,†
, Hailun Chen
1,3,†
, Zhenlin Jin
2
, Qiyang Zuo
1,3
, Yaohui Xu
1,3
and Kai He
1,3,
*
1
Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China;
wei.zhou1@siat.ac.cn (W.Z.); hl.chen@siat.ac.cn (H.C.); qy.zuo@siat.ac.cn (Q.Z.); yh.xu@siat.ac.cn (Y.X.)
2
School of Mechanical Engineering, Yanshan University, Qinhuangdao 066004, China; zljin@ysu.edu.cn
3
Shenzhen Key Laboratory of Precision Engineering, Shenzhen 518055, China
* Correspondence: kai.he@siat.ac.cn
† These authors contributed equally to this work.
Abstract:
It is too difficult to directly obtain the correspondence features between the two-dimensional
(2D) laser-range-finder (LRF) scan point and camera depth point cloud, which leads to a cumbersome
calibration process and low calibration accuracy. To address the problem, we propose a calibration
method to construct point-line constraint relations between 2D LRF and depth camera observational
features by using a specific calibration board. Through the observation of two different poses, we
construct the hyperstatic equations group based on point-line constraints and solve the coordinate
transformation parameters of 2D LRF and depth camera by the least square (LSQ) method. According
to the calibration error and threshold, the number of observation and the observation pose are
adjusted adaptively. After experimental verification and comparison with existing methods, the
method proposed in this paper easily and efficiently solves the problem of the joint calibration of the
2D LRF and depth camera, and well meets the application requirements of multi-sensor fusion for
mobile robots.
Keywords: two-dimensional laser-range-finder; depth camera; extrinsic calibration; data fusion
1. Introduction
With the rapid development of sensor technology and computer vision technology,
laser-range-finder (LRF) and cameras have become indispensable sensors for autonomous
driving, mobile robots and other fields [
1
]. Two-dimensional (2D) LRF is commonly used
to measure depth information in a single plane due to its high precision, light weight
and low power consumption. The camera acquires rich information, such as color and
texture, but it is sensitive to light and weather, resulting in its poor stability. On the other
hand, it is difficult for the camera to measure depth directly over long distances. Therefore,
laser vision fusion plays an important role in robot self-localization [
2
,
3
], environmental
perception [4], target tracking [5], and path planning [6].
To integrate data information from 2D LRF and depth cameras, the relative positional
relationship between the two sensors needs to be precisely known [
7
]. This is a classical
extrinsic calibration problem, where the objective is to determine the conversion relation-
ship between two coordinate systems. In contrast to 3D LRF, which identifies different
features, 2D LRF only measures depth information in a single plane, and it is difficult for
the camera to see the plane scanned by 2D LRF, which makes extrinsic calibration for 2D
LRF and cameras more challenging. Therefore, additional constraints must be used to find
the correspondence between the 2D LRF and the camera.
There has been a large amount of research work on the extrinsic calibration of 2D LRF
and cameras, which is divided into two categories: target-based calibration and non-target
calibration. References [
7
–
24
] are target-based calibration. Zhang and
Pless [8]
proposed
a method by using point constraints on a plane, but only two degrees of freedom are
Machines 2022, 10, 646. https://doi.org/10.3390/machines10080646 https://www.mdpi.com/journal/machines
