An image-based seismic damage assessment system

Abstract

Surveillance systems are commonly used by command centers to monitor and assess seismic damage inside buildings, such as schools, shopping malls, office buildings and skyscrapers. It is expected that checking camera images manually on monitors can be a very time-consuming and inefficient process, especially for a large surveillance system. One of the alternative ways is to first deploy sensors to monitor objects inside buildings, such as tables, cabinets, bookcases and so on, and, after fusing sensor data, to assess damage. However, deploying sensors can be impractical and costly when there are too many objects needed to be monitored. In this paper, we present IDEAS, an image-based disaster damage assessment system, to evaluate seismic damage inside buildings. IDEAS first compares images taken inside a building before and after an earthquake, it then maps the damage to a Mercalli intensity scale. In order to investigate the effectiveness and accuracy of IDEAS, we collect over forty pairs of closed-circuit television (CCTV) images from Youtube website. Each pair of images represents a real scenario of an earthquake inside a building. Our results show that IDEAS performs better than existing methods and can achieve an average accuracy of 97.6 % in mapping Mercalli intensity scale.