Building damage detection from post-earthquake aerial imagery using building grey-value and gradient orientation analyses

Abstract

The collapsed buildings due to 1999 Kocaeli earthquake were detected from post-event panchromatic aerial imagery based on grey-value and the gradient orientation of the buildings. The building boundaries were available and stored in a GIS as vector polygons. The building polygons were utilized to perform the assessments in a building specific manner. The spproach was implemented in a selected area of Golcuk, which is one of the urban areas most strongly hit by the earthquake. First, the buildings were selected one-by-one from the integrated vector (building boundaries) and raster (aerial photo) data set. The building damage detection process was then divided into two branches. In rhe first branch, the detection was performed using the building grey-value information, To do that, a greyvalue threshold (TI) \vas determined for discriminating the collapsed buildings from the un-collapsed ones. In the second branch, a group of operations including the gradient calculation and the determination of gradient orientation were performed. By utilizing the orientation information, an optimum threshold level (TZ) was determined for the standard deviation of the angle distribution of the building pixels. When assessing the condition of a building, the results of the two branches were combined and a final decision was made in an integrated manner. Of the 284 buildings analyzed, 254 were labeled correctly as coltapsed or un-collapsed providing an overall accuracy of 89.44%. The results reveal that the cotlapsed buildings due to the earthquake can be successfully detected from post-event aerial images.

Detection of collapsed buildings with the aerial images captured from UAV

“…When the threshold is too high (as 0.9), all the detection results will vanish. Figure 12(b) shows the effect of increasing the value of α during temporal tracking in (12). Through these experiments, it has been proved that increasing the value of α will erase the detection results that have low similarity among adjacent frames.…”

“…Other efforts [10][11][12] have been made to detect the collapsed buildings by only using the postseismic images. In [10], the feature points of collapsed buildings are extracted from the Airborne Laser Scanner (ALS) data with Hough Transformation, a binary detector of collapsed buildings is obtained by training the feature points of collapsed buildings and background component with the maximum entropy modeling method.…”

“…Despite its good results reported in that work, the performance of this work heavily depends on the manually selected thresholds, which means this work has to work under the supervision of humans. In [12], the features of damaged buildings are described by the gray image and a gradient orientation histogram. When the variation of histogram is over a threshold, an object is considered as a collapsed building, otherwise uncollapsed.…”

“…However, as the threshold keeps increasing, all the detection results will be erased gradually. (b) The effect of similarity α in (12). Although increasing the value of α could efficiently remove the false alarm, it will also remove some true positive detection.…”

Detecting earthquake damage levels using adaptive boosting

“…For this reason, lots of researches have been performed. Some of them have used different features such as shadow, edges and texture [3][4][5][6][7][8] and in some of them also have been applied machine learning approaches like neural networks, expert systems, support vector machine, k-nearest neighbors and baysian classification [9][10][11].…”