LARGE SCALE SEMANTIC SEGMENTATION OF VIRTUAL ENVIRONMENTS TO FACILITATE CORROSION MANAGEMENT

LARGE SCALE SEMANTIC SEGMENTATION OF VIRTUAL ENVIRONMENTS TO FACILITATE CORROSION MANAGEMENT

This paper reports the results of a study that aims to develop semi-automatic methods for assessing the degree of corrosion in industrial plant. We evaluated two fully convolutional networks (U-Net and DeepLab v3 +) to segment corroded areas in panoramic images of offshore platforms. The experimenta...

Full description

Bibliographic Details
Main Authors: R. L. Garcia, P. N. Happ, R. Q. Feitosa
Format: Article
Language:English
Published: Copernicus Publications 2021-06-01
Series:The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Online Access:https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLIII-B2-2021/465/2021/isprs-archives-XLIII-B2-2021-465-2021.pdf
id doaj-7e8ad470df5b4b58ad27149172bad8b0
record_format Article
spelling doaj-7e8ad470df5b4b58ad27149172bad8b02021-06-28T23:04:10ZengCopernicus PublicationsThe International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences1682-17502194-90342021-06-01XLIII-B2-202146547010.5194/isprs-archives-XLIII-B2-2021-465-2021LARGE SCALE SEMANTIC SEGMENTATION OF VIRTUAL ENVIRONMENTS TO FACILITATE CORROSION MANAGEMENTR. L. Garcia0P. N. Happ1R. Q. Feitosa2Pontifical Catholic University of Rio de Janeiro, Rio de Janeiro, BrazilPontifical Catholic University of Rio de Janeiro, Rio de Janeiro, BrazilPontifical Catholic University of Rio de Janeiro, Rio de Janeiro, BrazilThis paper reports the results of a study that aims to develop semi-automatic methods for assessing the degree of corrosion in industrial plant. We evaluated two fully convolutional networks (U-Net and DeepLab v3 +) to segment corroded areas in panoramic images of offshore platforms. The experimental analysis was based on two datasets built for this study. The datasets comprise 9,112 2D images and 3,732 panoramic images. Both FCNs trained on 2D images were tested on 2D images and cubic projections of panoramic images. In addition to pointing out encouraging results, the experiments indicated that most prediction errors concentrated in corrosion defects with a small pixel area.https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLIII-B2-2021/465/2021/isprs-archives-XLIII-B2-2021-465-2021.pdf
collection DOAJ
language English
format Article
sources DOAJ
author R. L. Garcia
P. N. Happ
R. Q. Feitosa
spellingShingle R. L. Garcia
P. N. Happ
R. Q. Feitosa
LARGE SCALE SEMANTIC SEGMENTATION OF VIRTUAL ENVIRONMENTS TO FACILITATE CORROSION MANAGEMENT
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
author_facet R. L. Garcia
P. N. Happ
R. Q. Feitosa
author_sort R. L. Garcia
title LARGE SCALE SEMANTIC SEGMENTATION OF VIRTUAL ENVIRONMENTS TO FACILITATE CORROSION MANAGEMENT
title_short LARGE SCALE SEMANTIC SEGMENTATION OF VIRTUAL ENVIRONMENTS TO FACILITATE CORROSION MANAGEMENT
title_full LARGE SCALE SEMANTIC SEGMENTATION OF VIRTUAL ENVIRONMENTS TO FACILITATE CORROSION MANAGEMENT
title_fullStr LARGE SCALE SEMANTIC SEGMENTATION OF VIRTUAL ENVIRONMENTS TO FACILITATE CORROSION MANAGEMENT
title_full_unstemmed LARGE SCALE SEMANTIC SEGMENTATION OF VIRTUAL ENVIRONMENTS TO FACILITATE CORROSION MANAGEMENT
title_sort large scale semantic segmentation of virtual environments to facilitate corrosion management
publisher Copernicus Publications
series The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
issn 1682-1750
2194-9034
publishDate 2021-06-01
description This paper reports the results of a study that aims to develop semi-automatic methods for assessing the degree of corrosion in industrial plant. We evaluated two fully convolutional networks (U-Net and DeepLab v3 +) to segment corroded areas in panoramic images of offshore platforms. The experimental analysis was based on two datasets built for this study. The datasets comprise 9,112 2D images and 3,732 panoramic images. Both FCNs trained on 2D images were tested on 2D images and cubic projections of panoramic images. In addition to pointing out encouraging results, the experiments indicated that most prediction errors concentrated in corrosion defects with a small pixel area.
url https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLIII-B2-2021/465/2021/isprs-archives-XLIII-B2-2021-465-2021.pdf
work_keys_str_mv AT rlgarcia largescalesemanticsegmentationofvirtualenvironmentstofacilitatecorrosionmanagement
AT pnhapp largescalesemanticsegmentationofvirtualenvironmentstofacilitatecorrosionmanagement
AT rqfeitosa largescalesemanticsegmentationofvirtualenvironmentstofacilitatecorrosionmanagement
_version_ 1721355772959391744

Similar Items