GIs and Image Understanding for Near-Real- Time Earthquake Damage Assessment

This paper evolved from the results of RADATT (RApid Damage Assessment Telematic Tool), a project funded by the European Commission, whose goal was to sensibly reduce the effects of a devastating seismic event by providing the responsible agencies a rapid and reliable damage detection and estimation of the extent and location of the suffered area. This result was accomplished by the integration, within a single user interface environment, of data access and standardization techniques, image processing tools, and GIS technology. A two-phase operating model has been conceived. In the pre-event era, images and data about buildings and infrastructures are collected and analyzed exploiting GIS capabilities. Immediately after the occurrence of the earthquake, the system must be ready to receive near-real-time imagery of the affected area to be compared with the pre-event data set. In this paper the capability of this quick change-detection analysis, given the availability of the pre-event information in the GIs environment, i s discussed. Introduction Recent natural disasters have stressed the importance of a rapid and reliable damage assessment and loss estimate in order to face both shortand long-term emergency response. It has been demonstrated that the entity of earthquake damage is only partly due to the strength of the natural event; it also depends on the delay and inadequacy of the rescue and reconstruction intervention. A reliable post-disaster damage assessment would allow the guarantee of rapid emergency response, and sensibly reduce the effects by providing an immediate estimate of the extent and location of the damaged area and making this knowledge available to the responsible agencies. Moreover, it should be pointed out that, in case of suddenly occurring natural disasters, the identification of damage by means of analytical models alone is often unreliable because soil conditions, geological complexities, and different environmental aspects may result in difficult to predict damage patterns. On the other hand, recent developments in image analysis techniques allow their integration with loss estimation and vulnerability models and make possible the investigation of more complex and effective approaches to the task of developing a reliable system for damage assessment. This paper outlines the principal issues for the development of a software system for damage and loss estimation in an emergency due to earthquake disasters. This goal is achieved by integrating, within a single user interface environment, existing database information, GIS (geographic information system) technology, and image processing techniques. First, a P. Gamba is with the Dipartimento di Elettronica, while F. Casciati is with the Dipartimento di Meccanica Strutturale, Universitg di Pavia, Via Ferrata, 1, 1-27100 Pavia, Italy. brief overview of the original RADATT project is offered, with particular stress on the image processing tools. A detailed damage assessment chain is then presented, followed by documentation of the experimental results obtained by applying the proposed approach to many case studies. Finally, conclusions are presented. The RADATT Project The RADATT (RApid Damage Assessment Telematic Tool) project (Casciati et al., 1996; Casciati and Giorgi, 1996a) discussed the feasibility for a system architecture satisfying the following requirements: Given that a seismic event occurred, can comparison of preand post-event imagery be used to detect the damage extent within a few hours from the shake? Can this comparison be conveniently supported by the appropriate interaction with an available building inventory? How is it more convenient to link, within a GIS environment, such a database with @e spatial geo-referenced information of interest? Thq project conceived a system architecture (Figure 1) and implemented modules to form a first prototype for seismic damage assessment and loss estimation from comparison of preand post-event images. The comparison is further