文章基本信息

标题：Structure from Motion of Underwater Scenes Considering Image Degradation and Refraction
本地全文：下载
作者：Xiaorui Qiao ; Yonghoon Ji ; Astushi Yamashita 等
期刊名称：IFAC PapersOnLine
印刷版ISSN：2405-8963
出版年度：2019
卷号：52
期号：22
页码：1-5
DOI：10.1016/j.ifacol.2019.11.051
语种：English
出版社：Elsevier
摘要：Structure from Motion (SfM) can reconstruct three-dimensional (3D) structuresusing only image sequences. However, SfM applied to the underwater environment is differentfrom the air environment because image formation in underwater environments suffers fromtwo major problems: image degradation and refraction.Consequently，images captured inunderwater environments are loss of contrast，hazy and distorted geometrically. To achieve anaccurate 3D reconstruction in such underwater environments,image degradation and refractionproblems should be solved simultaneously. In this work,we propose a systematic StM pipelinecontaining image enhancement and refraction reduced StM. The image enhancement partimproves the image quality for the following steps in the reconstrnction part. We confirm theeffectiveness of the proposed pipeline using images captured by a submerged robot in an extremeunderwater environment,the disaster area of Unit 3 Primary Containment Vessel (PCV) atFukushima Daiichi Nuclear Power Station. Experimental results show that our proposed methodcan successfully reconstruct the structures in that extreme underwater environment.
关键词：Underwater robots;3D Reconstruction;Structure from Motion;Image Enhancement;Refraction