SEMI-AUTOMATIC CITYSCAPE 3D MODEL RESTORATION USING GENERATIVE ADVERSARIAL NETWORK

SEMI-AUTOMATIC CITYSCAPE 3D MODEL RESTORATION USING GENERATIVE ADVERSARIAL NETWORK

The paper addresses the problem of a city heightmap restoration using satellite view image and some manually created area with 3D data. We propose the approach based on generative adversarial networks. Our algorithm contains three steps: low quality 3D restoration, buildings segmentation using resto...

Full description

Bibliographic Details
Main Authors: V. Gorbatsevich, B. Kulgildin, M. Melnichenko, O. Vygolov, Y. Vizilter
Format: Article
Language:English
Published: Copernicus Publications 2020-08-01
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-B2-2020/415/2020/isprs-archives-XLIII-B2-2020-415-2020.pdf
id doaj-92c5b7792ab642de9eefc0fb1cd67ba8
record_format Article
spelling doaj-92c5b7792ab642de9eefc0fb1cd67ba82020-11-25T03:38:28ZengCopernicus PublicationsThe International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences1682-17502194-90342020-08-01XLIII-B2-202041542010.5194/isprs-archives-XLIII-B2-2020-415-2020SEMI-AUTOMATIC CITYSCAPE 3D MODEL RESTORATION USING GENERATIVE ADVERSARIAL NETWORKV. Gorbatsevich0B. Kulgildin1M. Melnichenko2O. Vygolov3Y. Vizilter4Federal State Unitary Enterprise «State Research Institute of Aviation Systems», Russian FederationFederal State Unitary Enterprise «State Research Institute of Aviation Systems», Russian FederationFederal State Unitary Enterprise «State Research Institute of Aviation Systems», Russian FederationFederal State Unitary Enterprise «State Research Institute of Aviation Systems», Russian FederationFederal State Unitary Enterprise «State Research Institute of Aviation Systems», Russian FederationThe paper addresses the problem of a city heightmap restoration using satellite view image and some manually created area with 3D data. We propose the approach based on generative adversarial networks. Our algorithm contains three steps: low quality 3D restoration, buildings segmentation using restored model, and high-quality 3D restoration. CNN architecture based on original ResDilation blocks and ResNet is used for steps one and three. Training and test datasets were retrieved from National Lidar Dataset (United States) and the algorithm achieved approximately MSE = 3.84 m on this data. In addition, we tested our model on the completely different ISPRS Potsdam dataset and obtained MSE = 5.1 m.https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLIII-B2-2020/415/2020/isprs-archives-XLIII-B2-2020-415-2020.pdf
collection DOAJ
language English
format Article
sources DOAJ
author V. Gorbatsevich
B. Kulgildin
M. Melnichenko
O. Vygolov
Y. Vizilter
spellingShingle V. Gorbatsevich
B. Kulgildin
M. Melnichenko
O. Vygolov
Y. Vizilter
SEMI-AUTOMATIC CITYSCAPE 3D MODEL RESTORATION USING GENERATIVE ADVERSARIAL NETWORK
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
author_facet V. Gorbatsevich
B. Kulgildin
M. Melnichenko
O. Vygolov
Y. Vizilter
author_sort V. Gorbatsevich
title SEMI-AUTOMATIC CITYSCAPE 3D MODEL RESTORATION USING GENERATIVE ADVERSARIAL NETWORK
title_short SEMI-AUTOMATIC CITYSCAPE 3D MODEL RESTORATION USING GENERATIVE ADVERSARIAL NETWORK
title_full SEMI-AUTOMATIC CITYSCAPE 3D MODEL RESTORATION USING GENERATIVE ADVERSARIAL NETWORK
title_fullStr SEMI-AUTOMATIC CITYSCAPE 3D MODEL RESTORATION USING GENERATIVE ADVERSARIAL NETWORK
title_full_unstemmed SEMI-AUTOMATIC CITYSCAPE 3D MODEL RESTORATION USING GENERATIVE ADVERSARIAL NETWORK
title_sort semi-automatic cityscape 3d model restoration using generative adversarial network
publisher Copernicus Publications
series The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
issn 1682-1750
2194-9034
publishDate 2020-08-01
description The paper addresses the problem of a city heightmap restoration using satellite view image and some manually created area with 3D data. We propose the approach based on generative adversarial networks. Our algorithm contains three steps: low quality 3D restoration, buildings segmentation using restored model, and high-quality 3D restoration. CNN architecture based on original ResDilation blocks and ResNet is used for steps one and three. Training and test datasets were retrieved from National Lidar Dataset (United States) and the algorithm achieved approximately MSE = 3.84 m on this data. In addition, we tested our model on the completely different ISPRS Potsdam dataset and obtained MSE = 5.1 m.
url https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLIII-B2-2020/415/2020/isprs-archives-XLIII-B2-2020-415-2020.pdf
work_keys_str_mv AT vgorbatsevich semiautomaticcityscape3dmodelrestorationusinggenerativeadversarialnetwork
AT bkulgildin semiautomaticcityscape3dmodelrestorationusinggenerativeadversarialnetwork
AT mmelnichenko semiautomaticcityscape3dmodelrestorationusinggenerativeadversarialnetwork
AT ovygolov semiautomaticcityscape3dmodelrestorationusinggenerativeadversarialnetwork
AT yvizilter semiautomaticcityscape3dmodelrestorationusinggenerativeadversarialnetwork
_version_ 1724542215150108672

Similar Items