Abstract. Many people have got injured and died by sudden disasters such as fires and terrorisms. We have proposed an Emergency Rescue Support System (ERESS) for reducing victims at the time of disaster. ERESS operates under mobile ad-hoc networks (MANET) composed of handheld terminals such as smartphones and tablets. ERESS terminals have disaster detection algorithm and plural sensors such as acceleration, angular velocity, and geomagnetism. ERESS detects the disaster from the behavior analysis of people by the sensors. In this paper, we propose a new disaster detection method by performing the machine learning in the group using a support vector domain description (SVDD). We are able to detect the abnormal behavior of people by using this method. The results of disaster simulation experiments show the validity of the proposed method.
A lot of people have got injured and died by sudden disasters such as fires and terrorisms. We have proposed an Emergency Rescue Support System (ERESS) for the purpose of reducing victims at the time of disaster. This system operates under mobile ad-hoc networks (MANET). So ERESS uses handheld terminals (ERESS terminals) such as smartphones and tablets. The ERESS terminals have disaster detection algorithm and plural sensors (acceleration, angular velocity, and geomagnetism). The sensors are used for the behavior analysis of ERESS terminal holders. By using the results of the analysis, the system detects the disaster from the behavior of people. In this paper, we propose a new disaster detection method by performing the machine learning in the group using a support vector domain description (SVDD). We are able to detect the behavior that is different from normal state of people in the disaster by using this method. The results of the performance evaluation by disaster simulation experiments show the validity of the proposed method.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.