A Deep Learning Model for Road Damage Detection After an Earthquake Based on Synthetic Aperture Radar (SAR) and Field Datasets

Sadra Karimzadeh; Mohammad Ghasemi; Masashi Matsuoka; Koichi Yagi; Abdullah Can Zulfikar

doi:10.1109/JSTARS.2022.3189875

A Deep Learning Model for Road Damage Detection After an Earthquake Based on Synthetic Aperture Radar (SAR) and Field Datasets

Sadra Karimzadeh^*, Mohammad Ghasemi, Masashi Matsuoka, Koichi Yagi, Abdullah Can Zulfikar

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

17 Citations (Scopus)

Abstract

This article is a new assessment of damaged roads after the Kumamoto earthquake in southern Japan (2016) using remotely sensed synthetic aperture radar (SAR) data, field data and deep learning. Three SAR images from descending orbits of Sentinel-1 in vertical-vertical polarizations are considered for radiometric calibration, geocoding and interferometric analyses. Field data in terms of the international roughness index (IRI) were gathered over more than 530 km using a smartphone accelerometer and the BumpRecorder application. The relationship between SAR data and IRI data was investigated in a binary (0 and 1) mode to establish a multilayer perceptron model of damaged and intact roads. We found the remote sensing SAR datasets suitable, not only for the detection of damaged roads, but also as an indicator of road roughness changes. The classification results for damaged and intact roads indicated that our datasets (SAR and field measurements), together with a deep learning model, yielded acceptable overall accuracy (87.1%).

Original language	English
Pages (from-to)	5753-5765
Number of pages	13
Journal	IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Volume	15
DOIs	https://doi.org/10.1109/JSTARS.2022.3189875
Publication status	Published - 2022
Externally published	Yes

Bibliographical note

Publisher Copyright:
© 2008-2012 IEEE.

Funding

This work was supported by the Japanese Society for the Promotion of Science (JSPS) Grants-in-Aid for Scientific Research (KAKENHI) under Grant 20H02411.

Funders	Funder number
Japan Society for the Promotion of Science	20H02411

Keywords

Deep learning
Kumamoto
international roughness index (IRI)
synthetic aperture radar (SAR)

Access to Document

10.1109/JSTARS.2022.3189875

Cite this

@article{6b5a46f499634c98ac5825ac5c3524d3,

title = "A Deep Learning Model for Road Damage Detection After an Earthquake Based on Synthetic Aperture Radar (SAR) and Field Datasets",

abstract = "This article is a new assessment of damaged roads after the Kumamoto earthquake in southern Japan (2016) using remotely sensed synthetic aperture radar (SAR) data, field data and deep learning. Three SAR images from descending orbits of Sentinel-1 in vertical-vertical polarizations are considered for radiometric calibration, geocoding and interferometric analyses. Field data in terms of the international roughness index (IRI) were gathered over more than 530 km using a smartphone accelerometer and the BumpRecorder application. The relationship between SAR data and IRI data was investigated in a binary (0 and 1) mode to establish a multilayer perceptron model of damaged and intact roads. We found the remote sensing SAR datasets suitable, not only for the detection of damaged roads, but also as an indicator of road roughness changes. The classification results for damaged and intact roads indicated that our datasets (SAR and field measurements), together with a deep learning model, yielded acceptable overall accuracy (87.1%).",

keywords = "Deep learning, Kumamoto, international roughness index (IRI), synthetic aperture radar (SAR)",

author = "Sadra Karimzadeh and Mohammad Ghasemi and Masashi Matsuoka and Koichi Yagi and Zulfikar, {Abdullah Can}",

note = "Publisher Copyright: {\textcopyright} 2008-2012 IEEE.",

year = "2022",

doi = "10.1109/JSTARS.2022.3189875",

language = "English",

volume = "15",

pages = "5753--5765",

journal = "IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing",

issn = "1939-1404",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

A Deep Learning Model for Road Damage Detection After an Earthquake Based on Synthetic Aperture Radar (SAR) and Field Datasets. / Karimzadeh, Sadra; Ghasemi, Mohammad; Matsuoka, Masashi et al.
In: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, Vol. 15, 2022, p. 5753-5765.

Research output: Contribution to journal › Article › peer-review

TY - JOUR

T1 - A Deep Learning Model for Road Damage Detection After an Earthquake Based on Synthetic Aperture Radar (SAR) and Field Datasets

AU - Karimzadeh, Sadra

AU - Ghasemi, Mohammad

AU - Matsuoka, Masashi

AU - Yagi, Koichi

AU - Zulfikar, Abdullah Can

PY - 2022

Y1 - 2022

N2 - This article is a new assessment of damaged roads after the Kumamoto earthquake in southern Japan (2016) using remotely sensed synthetic aperture radar (SAR) data, field data and deep learning. Three SAR images from descending orbits of Sentinel-1 in vertical-vertical polarizations are considered for radiometric calibration, geocoding and interferometric analyses. Field data in terms of the international roughness index (IRI) were gathered over more than 530 km using a smartphone accelerometer and the BumpRecorder application. The relationship between SAR data and IRI data was investigated in a binary (0 and 1) mode to establish a multilayer perceptron model of damaged and intact roads. We found the remote sensing SAR datasets suitable, not only for the detection of damaged roads, but also as an indicator of road roughness changes. The classification results for damaged and intact roads indicated that our datasets (SAR and field measurements), together with a deep learning model, yielded acceptable overall accuracy (87.1%).

AB - This article is a new assessment of damaged roads after the Kumamoto earthquake in southern Japan (2016) using remotely sensed synthetic aperture radar (SAR) data, field data and deep learning. Three SAR images from descending orbits of Sentinel-1 in vertical-vertical polarizations are considered for radiometric calibration, geocoding and interferometric analyses. Field data in terms of the international roughness index (IRI) were gathered over more than 530 km using a smartphone accelerometer and the BumpRecorder application. The relationship between SAR data and IRI data was investigated in a binary (0 and 1) mode to establish a multilayer perceptron model of damaged and intact roads. We found the remote sensing SAR datasets suitable, not only for the detection of damaged roads, but also as an indicator of road roughness changes. The classification results for damaged and intact roads indicated that our datasets (SAR and field measurements), together with a deep learning model, yielded acceptable overall accuracy (87.1%).

KW - Deep learning

KW - Kumamoto

KW - international roughness index (IRI)

KW - synthetic aperture radar (SAR)

UR - http://www.scopus.com/inward/record.url?scp=85134296823&partnerID=8YFLogxK

U2 - 10.1109/JSTARS.2022.3189875

DO - 10.1109/JSTARS.2022.3189875

M3 - Article

AN - SCOPUS:85134296823

SN - 1939-1404

VL - 15

SP - 5753

EP - 5765

JO - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

JF - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

ER -

A Deep Learning Model for Road Damage Detection After an Earthquake Based on Synthetic Aperture Radar (SAR) and Field Datasets

Abstract

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this