Back to search

Michigan State University

LIDAR AND CAMERA CALIBRATION USING A MOUNTED SPHERE

Abstract

dc:description

Extrinsic calibration between lidar and camera sensors is needed for multi-modal sensor data fusion. However, obtaining precise extrinsic calibration can be tedious, computationally expensive, or involve elaborate apparatus. This thesis proposes a simple, fast, and robust method performing extrinsic calibration between a camera and lidar. The only required calibration target is a hand-held colored sphere mounted on a whiteboard. The convolutional neural networks are developed to automatically localize the sphere relative to the camera and the lidar. Then using the localization covariance models, the relative pose between the camera and lidar is derived. To evaluate the accuracy of our method, we record image and lidar data of a sphere at a set of known grid positions by using two rails mounted on a wall. The accurate calibration results are demonstrated by projecting the grid centers into the camera image plane and finding the error between these points and the hand-labeled sphere centers.

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Li, Jiajia
Contributors dc:contributor
  • Morris, Daniel
  • Radha, Hayder
  • Li, Zhaojian

Subjects

dc:subject × 2

Rights

dc:rights
Statement dc:rights
  • In Copyright
Language dc:language
English

Identifiers

dc:identifier.*
Identifier
etd:49462
local:Li_grad.msu_0128N_17951

Chain of custody

source
Harvested from
Michigan State University
Base URL
d.lib.msu.edu/oai
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Li, Jiajia. LIDAR AND CAMERA CALIBRATION USING A MOUNTED SPHERE. 2020. https://doi.org/doi:10.25335/pg5g-7a22