Multi-Hazard and Spatial Transferability of a CNN for Automated Building Damage Assessment

Abstract

Automated classification of building damage in remote sensing images enables the rapid and spatially extensive assessment of the impact of natural hazards, thus speeding up emergency response efforts. Convolutional neural networks (CNNs) can reach good performance on such a task in experimental settings. How CNNs perform when applied under operational emergency conditions, with unseen data and time constraints, is not well studied. This study focuses on the applicability of a CNN-based model in such scenarios. We performed experiments on 13 disasters that differ in natural hazard type, geographical location, and image parameters. The types of natural hazards were hurricanes, tornadoes, floods, tsunamis, and volcanic eruptions, which struck across North America, Central America, and Asia. We used 175,289 buildings from the xBD dataset, which contains human-annotated multiclass damage labels on high-resolution satellite imagery with red, green, and blue (RGB) bands. First, our experiments showed that the performance in terms of area under the curve does not correlate with the type of natural hazard, geographical region, and satellite parameters such as the off-nadir angle. Second, while performance differed highly between occurrences of disasters, our model still reached a high level of performance without using any labeled data of the test disaster during training. This provides the first evidence that such a model can be effectively applied under operational conditions, where labeled damage data of the disaster cannot be available timely and thus model (re-)training is not an option.

Transferability of Convolutional Neural Network Models for Identifying Damaged Buildings Due to Earthquake

“…The xBD dataset is a large-scale disaster dataset, containing eight types of disasters, including earthquakes, tsunamis, floods, landslides, wind, volcanic eruption, wildfire, and dam collapse. It contains pre-disaster and post-disaster RGB satellite images (e.g., Worldview, Quickbird, GeoEye) of 19 disaster events with a resolution equal to or less than 0.8 m [52,57]. The dataset also contains building polygons developed with the pre-disaster images and damage grade interpreted from post-disaster images.…”

“…Deep learning technology has recently been widely used in VHR remote sensing image applications [43][44][45]. Lots of explorations in disaster assessments were performed in the literature [22,23,[25][26][27][28][29]42,[46][47][48][49][50][51][52][53][54]. Convolutional Neural Network (CNN) is a deep learning technique that can automatically learn the most effective features from samples while training.…”

“…As the acquisition of high-resolution disaster images is very expensive, the research on integrating multiple disaster types is very limited. The xBD disaster public dataset provides strong support for the exploration in this field [46,49,52,53]. In one of the latest applications related to xBD, five disaster types (i.e., hurricane, tornado, flood, tsunami, and volcanic eruption) were chosen to explore the applicability of the models trained by different combinations of input building samples [52].…”

“…The xBD disaster public dataset provides strong support for the exploration in this field [46,49,52,53]. In one of the latest applications related to xBD, five disaster types (i.e., hurricane, tornado, flood, tsunami, and volcanic eruption) were chosen to explore the applicability of the models trained by different combinations of input building samples [52]. The results indicate that the accuracy of the CNN model is independent of the geographic areas and satellite parameters (e.g., off-nadir angle).…”

A Machine Learning Framework for Assessing Seismic Hazard Safety of Reinforced Concrete Buildings

“…A substantial literature has been published in the attempts of integrating several methods from various domains with RVS, for instance, statistical methods [11,12], Artificial Neural Network (ANN) [13][14][15][16][17], multi-criteria decision making [18,19], and type-1 [20][21][22][23] and type-2 [24,25] fuzzy logic systems are frequently assimilated within RVS for increasing the interface and efficacy of seismic vulnerability screening. However, there are methods developed to evaluate the damage and change detection of buildings by remote sensing and image analysis [26,27]. The parameter significance and the selection of the optimum number of parameters have been used effectively by Morfidis and Kostinakis [28].…”