Detection of Collapsed Bridges from Multi-Temporal SAR Intensity Images by Machine Learning Techniques

• Published in: Remote sensing (Basel, Switzerland) Vol. 13; no. 17; p. 3508
• Main Authors: Liu, Wen, Maruyama, Yoshihisa, Yamazaki, Fumio
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01-09-2021
• Subjects: Accuracy; Backscattering; Bridge failure; Bridges; Damage assessment; Datasets; Deep learning; Disasters; Earthquake damage; Earthquakes; Emergency preparedness; Emergency response; Feature extraction; Floods; Image acquisition; Image resolution; Learning algorithms; logistic regression; Machine learning; multi-temporal SAR intensity images; Oversampling; random forest; Records & achievements; Remote sensing; Rivers; Roads; Seismic activity; Seismic engineering; Sensors; Storm damage; Support vector machines; Synthetic aperture radar; the 2011 Tohoku-oki earthquake; the 2020 July floods in Japan; Transportation networks; Tsunamis; Japan
• ISSN: 2072-4292; 2072-4292
• DOI: 10.3390/rs13173508

Bridges are an important part of road networks in an emergency period, as well as in ordinary times. Bridge collapses have occurred as a result of many recent disasters. Synthetic aperture radar (SAR), which can acquire images under any weather or sunlight conditions, has been shown to be effective in assessing the damage situation of structures in the emergency response phase. We investigate the backscattering characteristics of washed-away or collapsed bridges from the multi-temporal high-resolution SAR intensity imagery introduced in our previous studies. In this study, we address the challenge of building a model to identify collapsed bridges using five change features obtained from multi-temporal SAR intensity images. Forty-four bridges affected by the 2011 Tohoku-oki earthquake, in Japan, and forty-four bridges affected by the 2020 July floods, also in Japan, including a total of 21 collapsed bridges, were divided into training, test, and validation sets. Twelve models were trained, using different numbers of features as input in random forest and logistic regression methods. Comparing the accuracies of the validation sets, the random forest model trained with the two mixed events using all the features showed the highest capability to extract collapsed bridges. After improvement by introducing an oversampling technique, the F-score for collapsed bridges was 0.87 and the kappa coefficient was 0.82, showing highly accurate agreement.