Iterative refinement for underwater 3D reconstruction: application to disposed underwater munitions in the Baltic Sea.

Song, Yifan, Köser, Kevin, Kwasnitschka, Tom and Koch, Reinhard (2019) Iterative refinement for underwater 3D reconstruction: application to disposed underwater munitions in the Baltic Sea. Open Access ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII-2/W10 . pp. 181-187. DOI 10.5194/isprs-archives-XLII-2-W10-181-2019.

[thumbnail of isprs-archives-XLII-2-W10-181-2019.pdf]
Preview
Text
isprs-archives-XLII-2-W10-181-2019.pdf - Published Version
Available under License Creative Commons: Attribution 4.0.

Download (1MB) | Preview

Supplementary data:

Abstract

With the rapid development and availability of underwater imaging technologies, underwater visual recording is widely used for a variety of tasks. However, quantitative imaging and photogrammetry in the underwater case has a lot of challenges (strong geometry distortion and radiometry issues) that limit the traditional photogrammetric workflow in underwater applications. This paper presents an iterative refinement approach to cope with refraction induced distortion while building on top of a standard photogrammetry pipeline. The approach uses approximate geometry to compensate for water refraction effects in images and then brings the new images into the next iteration of 3D reconstruction until the update of resulting depth maps becomes neglectable. Afterwards, the corrected depth map can also be used to compensate the attenuation effect in order to get a more realistic color for the 3D model. To verify the geometry improvement of the proposed approach, a set of images with air-water refraction effect were rendered from a ground truth model and the iterative refinement approach was applied to improve the 3D reconstruction. At the end, this paper also shows its application results for 3D reconstruction of a dump site for underwater munition in the Baltic Sea for which a visual monitoring approach is desired.

Document Type: Article
Keywords: Underwater Photogrammetry, 3D Reconstruction, Refraction Correction, Underwater Image Restoration, Underwater Munitions
Research affiliation: OceanRep > GEOMAR > FB2 Marine Biogeochemistry > FB2-MG Marine Geosystems
OceanRep > GEOMAR > FB2 Marine Biogeochemistry > FB2-MG Marine Geosystems > DeepSea Monitoring
OceanRep > GEOMAR > FB4 Dynamics of the Ocean Floor > FB4-MUHS
Kiel University
Refereed: Yes
Open Access Journal?: Yes
Publisher: Internationale Gesellschaft für Photogrammetrie und Fernerkundung
Date Deposited: 30 Apr 2019 08:15
Last Modified: 31 Jan 2022 09:13
URI: https://oceanrep.geomar.de/id/eprint/46417

Actions (login required)

View Item View Item