期刊名称：International Journal of Security and Its Applications
印刷版ISSN：1738-9976
出版年度：2015
卷号：9
期号：7
页码：59-68
DOI：10.14257/ijsia.2015.9.7.05
出版社：SERSC
摘要：Because of environmental issues such as global warming, natural disaster is a growing trend. Accordingly, interest in minimizing structural damage due to natural disasters is also increasing. One method for reducing these problems of disaster damage is to detect problematic situations quickly using Closed-circuit televisions (CCTVs), which are image sensors that allow damage to be controlled and countermeasures to be prepared. In the case of structural deformation situations occurring due to natural disasters, deformations tend to continue until the structure is recovered. However, since intelligent CCTVs have thus far failed to recognize or track moving objects, these CCTVs are not suitable for detecting the maintenance of structure deformation. Therefore, in the present study, a technique will be presented through which structural situations suspected of deformation (i.e., situations in which problems will occur if the current state is maintained) can be intelligently detected using difference images based on video images taken from actual structures at fixed positions.
