GeoRisk Journal article

This paper contains two main parts. First, we give an overview of photogrammetric and remote sensing techniques in applications related to disaster management procedures. Remotely sensed data potentially provides valuable seismicity-related data in terms of both assessing the potential of risk and defining the state of vulnerability. The paper illustrates the use of advanced remote sensing technologies and methods for urban inventory and damage assessment. Secondly, we explain our proposed method for automatic damage assessment using multi view aerial images. We document the use of a novel methodology for assessing the damages of man-made objects. The problem is formulated in general, and for validation the approach is applied to a study area. This area is a part of the city of Bam, Iran that was hit by a strong earthquake on December 26, 2003. For our purpose digital surface models (DSMs) were created automatically from preand post-earthquake aerial images. We have explored the use of different kinds of extracted features, defined in both object and image spaces. This method utilizes a methodology of applying Bayesian networks to a multiview and multi-modal damaged objects description. By comparing preand post-earthquake data the amount of damage can be assessed. Based on visual inspection of the stereo images, three levels of damage (total collapse, partial collapse, no damage) were considered. The overall success rate – total number of correctly classified objects divided by the total number of samples – was found to be 80.7%. The results of the analysis show that a good degree of agreement has been reached between the assessment results and the reference data. The suggested approach is promising, but there is room for improvement.