Characteristics of texture index of damaged buildings using time-series high-resolution satellite images for the 2024 Noto Peninsula earthquake
Masashi SONOBE

Department of Civil Engineering College of Science and Technology, Nihon University

Abstract

In the Noto Peninsula earthquake that occurred in January 2024, the earthquake and tsunami caused human casualties, collapsed buildings, and caused severe damage to infrastructure. In cases where damage is severe and widespread, it is necessary to grasp the extent of the damage as quickly as possible, and collecting damage information using satellite remote sensing is effective. Furthermore, the resolution of the satellite image has been increased, and Method of object-based and pixel-based texture for extracting detailed damage information of the building and grasping the damaged area have been studied. However, there have been few time-series surveys of damage and recovery before and after a disaster and one year after the disaster using texture analysis of satellite images. In this paper, the distribution characteristics of texture indices obtained from high-resolution satellite images observed immediately after the earthquake and approximately one year after the earthquake were evaluated using detailed field survey results to investigate building damage and recovery status before and after the disaster. The target area was the Horyu-cho area of Suzu City, where buildings were damaged by the earthquake and tsunami caused by the 2024 Noto Peninsula earthquake. First, histogram matching was performed using satellite images taken three times before and after the disaster, and then texture indices were calculated using a co-occurrence matrix to evaluate the distribution. In this study, we used dissimilarity and homogeneity as a texture index based on the Gray-Level Co-occurrence Matrix(GLCM). The band used was the red band after pan-sharpening. The pixel size was resampled to 0.5m. The frequency distribution of pixel values of usage data is different. For this reason, pixel values with cumulative frequency between 2% and 98% were used to convert to 8-bit images. First, the characteristics were grasped from the obtained texture index by visual interpretation. In addition, the average value within the building polygon was calculated, and the characteristics of the texture indices of damaged and restored buildings immediately after the disaster, depending on whether or not there were flooded areas, were grasped from the results of field surveys.
The results showed that flooded areas tend to be more dissimilarity and less homogeneity immediately after a disaster. These were presumably the result of buildings being washed away or collapsed in areas flooded by the tsunami. This tendency is smaller in non-flooded areas, but the dissimilarity tends to be higher immediately after a disaster. This was presumably due to debris from buildings that collapsed due to the earthquake in non-flooded areas. In addition, damage levels were divided according to whether or not there was flooding and whether or not buildings collapsed, and the results were compared with the index values, suggesting the possibility of classifying runoff and building collapse. In addition, damage levels were divided according to whether or not there was flooding and whether or not buildings collapsed, and the results were compared with the index values, suggesting the possibility of classifying runoff and building collapse. From these results, we confirmed the effectiveness of texture analysis using high-resolution satellite images in grasping the damaged buildings before and immediately after the disaster and in the restoration situation one year after the disaster. This suggests that the method can be applied to disasters that require early collection of damage information in the event of a disaster.

Keywords: Disaster, Earthquake, Tsunami, Damaged buildings, Texture analysis

Topic: Topic A: General Remote Sensing