Extraction of buildings in vhr sar images using fully convolution neural networks

Muhammad Shahzad, Michael Maurer, Friedrich Fraundorfer, Yuanyuan Wang, Xiao Xiang Zhu

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

Modern spaceborne synthetic aperture radar (SAR) sensors, such as TerraSAR-X/TanDEM-X and COSMO-SkyMed, can deliver very high resolution (VHR) data beyond the inherent spatial scales (on the order of 1m) of buildings, constituting invaluable data source for large-scale urban mapping. Processing this VHR data with advanced interferometric techniques, such as SAR tomography (TomoSAR), enables the generation of 3-D (or even 4-D) TomoSAR point clouds from space. In this paper, we present a novel and generic workflow that exploits these TomoSAR point clouds in a way that is capable to automatically produce benchmark annotated (buildings/nonbuildings) SAR datasets. These annotated datasets (building masks) have been utilized to construct and train the state-ofthe- A rt deep Fully Convolution Neural Networks with an additional Conditional Random Field represented as a Recurrent Neural Network to detect building regions in a single VHR SAR image. The results of building detection are illustrated and validated over TerraSAR-X VHR spotlight SAR image covering approximately 39 km2 . almost the whole city of Berlin . with mean pixel accuracies of around 93.84%.

Original languageEnglish
Title of host publication2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers
Pages4367-4370
Number of pages4
ISBN (Electronic)9781538671504
DOIs
Publication statusPublished - 31 Oct 2018
Event38th Annual IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Valencia, Spain
Duration: 22 Jul 201827 Jul 2018

Publication series

NameInternational Geoscience and Remote Sensing Symposium (IGARSS)
Volume2018-July

Conference

Conference38th Annual IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018
CountrySpain
CityValencia
Period22/07/1827/07/18

Fingerprint

Synthetic aperture radar
Convolution
synthetic aperture radar
Neural networks
TerraSAR-X
TanDEM-X
COSMO-SkyMed
Recurrent neural networks
tomography
train
Tomography
Masks
pixel
Pixels
sensor
Sensors
Processing

Keywords

  • Building Detection
  • Fully Convolution Neural Networks
  • OpenStreetMap
  • SAR Tomography
  • Synthetic Aperture Radar (SAR)
  • TerraSAR-X/TanDEM-X

ASJC Scopus subject areas

  • Computer Science Applications
  • Earth and Planetary Sciences(all)

Cite this

Shahzad, M., Maurer, M., Fraundorfer, F., Wang, Y., & Zhu, X. X. (2018). Extraction of buildings in vhr sar images using fully convolution neural networks. In 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings (pp. 4367-4370). [8519603] (International Geoscience and Remote Sensing Symposium (IGARSS); Vol. 2018-July). Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/IGARSS.2018.8519603

Extraction of buildings in vhr sar images using fully convolution neural networks. / Shahzad, Muhammad; Maurer, Michael; Fraundorfer, Friedrich; Wang, Yuanyuan; Zhu, Xiao Xiang.

2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings. Institute of Electrical and Electronics Engineers, 2018. p. 4367-4370 8519603 (International Geoscience and Remote Sensing Symposium (IGARSS); Vol. 2018-July).

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Shahzad, M, Maurer, M, Fraundorfer, F, Wang, Y & Zhu, XX 2018, Extraction of buildings in vhr sar images using fully convolution neural networks. in 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings., 8519603, International Geoscience and Remote Sensing Symposium (IGARSS), vol. 2018-July, Institute of Electrical and Electronics Engineers, pp. 4367-4370, 38th Annual IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018, Valencia, Spain, 22/07/18. https://doi.org/10.1109/IGARSS.2018.8519603
Shahzad M, Maurer M, Fraundorfer F, Wang Y, Zhu XX. Extraction of buildings in vhr sar images using fully convolution neural networks. In 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings. Institute of Electrical and Electronics Engineers. 2018. p. 4367-4370. 8519603. (International Geoscience and Remote Sensing Symposium (IGARSS)). https://doi.org/10.1109/IGARSS.2018.8519603
Shahzad, Muhammad ; Maurer, Michael ; Fraundorfer, Friedrich ; Wang, Yuanyuan ; Zhu, Xiao Xiang. / Extraction of buildings in vhr sar images using fully convolution neural networks. 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings. Institute of Electrical and Electronics Engineers, 2018. pp. 4367-4370 (International Geoscience and Remote Sensing Symposium (IGARSS)).
@inproceedings{2c7bb06bb4804149b2d64a20b1aa07ec,
title = "Extraction of buildings in vhr sar images using fully convolution neural networks",
abstract = "Modern spaceborne synthetic aperture radar (SAR) sensors, such as TerraSAR-X/TanDEM-X and COSMO-SkyMed, can deliver very high resolution (VHR) data beyond the inherent spatial scales (on the order of 1m) of buildings, constituting invaluable data source for large-scale urban mapping. Processing this VHR data with advanced interferometric techniques, such as SAR tomography (TomoSAR), enables the generation of 3-D (or even 4-D) TomoSAR point clouds from space. In this paper, we present a novel and generic workflow that exploits these TomoSAR point clouds in a way that is capable to automatically produce benchmark annotated (buildings/nonbuildings) SAR datasets. These annotated datasets (building masks) have been utilized to construct and train the state-ofthe- A rt deep Fully Convolution Neural Networks with an additional Conditional Random Field represented as a Recurrent Neural Network to detect building regions in a single VHR SAR image. The results of building detection are illustrated and validated over TerraSAR-X VHR spotlight SAR image covering approximately 39 km2 . almost the whole city of Berlin . with mean pixel accuracies of around 93.84{\%}.",
keywords = "Building Detection, Fully Convolution Neural Networks, OpenStreetMap, SAR Tomography, Synthetic Aperture Radar (SAR), TerraSAR-X/TanDEM-X",
author = "Muhammad Shahzad and Michael Maurer and Friedrich Fraundorfer and Yuanyuan Wang and Zhu, {Xiao Xiang}",
year = "2018",
month = "10",
day = "31",
doi = "10.1109/IGARSS.2018.8519603",
language = "English",
series = "International Geoscience and Remote Sensing Symposium (IGARSS)",
publisher = "Institute of Electrical and Electronics Engineers",
pages = "4367--4370",
booktitle = "2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings",
address = "United States",

}

TY - GEN

T1 - Extraction of buildings in vhr sar images using fully convolution neural networks

AU - Shahzad, Muhammad

AU - Maurer, Michael

AU - Fraundorfer, Friedrich

AU - Wang, Yuanyuan

AU - Zhu, Xiao Xiang

PY - 2018/10/31

Y1 - 2018/10/31

N2 - Modern spaceborne synthetic aperture radar (SAR) sensors, such as TerraSAR-X/TanDEM-X and COSMO-SkyMed, can deliver very high resolution (VHR) data beyond the inherent spatial scales (on the order of 1m) of buildings, constituting invaluable data source for large-scale urban mapping. Processing this VHR data with advanced interferometric techniques, such as SAR tomography (TomoSAR), enables the generation of 3-D (or even 4-D) TomoSAR point clouds from space. In this paper, we present a novel and generic workflow that exploits these TomoSAR point clouds in a way that is capable to automatically produce benchmark annotated (buildings/nonbuildings) SAR datasets. These annotated datasets (building masks) have been utilized to construct and train the state-ofthe- A rt deep Fully Convolution Neural Networks with an additional Conditional Random Field represented as a Recurrent Neural Network to detect building regions in a single VHR SAR image. The results of building detection are illustrated and validated over TerraSAR-X VHR spotlight SAR image covering approximately 39 km2 . almost the whole city of Berlin . with mean pixel accuracies of around 93.84%.

AB - Modern spaceborne synthetic aperture radar (SAR) sensors, such as TerraSAR-X/TanDEM-X and COSMO-SkyMed, can deliver very high resolution (VHR) data beyond the inherent spatial scales (on the order of 1m) of buildings, constituting invaluable data source for large-scale urban mapping. Processing this VHR data with advanced interferometric techniques, such as SAR tomography (TomoSAR), enables the generation of 3-D (or even 4-D) TomoSAR point clouds from space. In this paper, we present a novel and generic workflow that exploits these TomoSAR point clouds in a way that is capable to automatically produce benchmark annotated (buildings/nonbuildings) SAR datasets. These annotated datasets (building masks) have been utilized to construct and train the state-ofthe- A rt deep Fully Convolution Neural Networks with an additional Conditional Random Field represented as a Recurrent Neural Network to detect building regions in a single VHR SAR image. The results of building detection are illustrated and validated over TerraSAR-X VHR spotlight SAR image covering approximately 39 km2 . almost the whole city of Berlin . with mean pixel accuracies of around 93.84%.

KW - Building Detection

KW - Fully Convolution Neural Networks

KW - OpenStreetMap

KW - SAR Tomography

KW - Synthetic Aperture Radar (SAR)

KW - TerraSAR-X/TanDEM-X

UR - http://www.scopus.com/inward/record.url?scp=85060662337&partnerID=8YFLogxK

U2 - 10.1109/IGARSS.2018.8519603

DO - 10.1109/IGARSS.2018.8519603

M3 - Conference contribution

T3 - International Geoscience and Remote Sensing Symposium (IGARSS)

SP - 4367

EP - 4370

BT - 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings

PB - Institute of Electrical and Electronics Engineers

ER -