Auto-calibration of line-laser structured-light seafloor mapping systems

Abstract

The sub-centimetre resolution of line-laser generated maps exceeds acoustic bathymetry and provides detailed information needed for inspection and ecological research, but makes the maps more sensitive to errors stemming from (navigation and) mapping sensor calibration. The high resolution is achieved using a large relative baseline between the camera and laser apparatus. Their extrinsic parameters are often incidentally changed between missions when the vehicle they are attached to undergoes maintenance, making it desirable to have a means of calibrating these parameters in-situ and using natural scenes. Existing methods require the use of a second camera, which increases the complexity of the mapping hardware and often requires dedicated calibration manoeuvres to be performed. In this paper we present a novel in-situ calibration method using natural scenes that does not require the use of a second camera, but instead uses the ordinary data captured during a survey mission to perform a non-linear least squares optimisation that reduces inconsistencies between overlapping measurements. We take advantage of a dual line-laser setup to generate continuously overlapping measurements for calibration that have inherently small uncertainty in their relative pose estimates and demonstrate reduced inconsistency between overlapping data using our optimised calibration parameters over that of CAD-derived values.

More information

Identifiers

ORCID for David Stanley: orcid.org/0000-0003-3699-3203

ORCID for Adrian Bodenmann: orcid.org/0000-0002-3195-0602

ORCID for Miguel Massot Campos: orcid.org/0000-0002-1202-0362

Catalogue record

Export record

Altmetrics