Research On The Detection Method Of Abnormal Behavior Of The Elderly Based On Time Law

It points out that the number of members of the family has been on the decrease and the demand in the elderly care service is becoming more important in the 2015 Chinese Family Development Report. Because of a variety of disease of the elderly which are the important monitored people, there are a lot of danger may occur in their daily lives. So, a method which can be proposed for automatically detecting the living-alone elderly's abnormal behavior is important.This paper presents some research in the elderly's activities in their daily lives when they live alone in monitored areas.· To get the rea time location of the elderly when they begin to move in their monitored areas with the real-time location system based on the UWB (Ultra-wideband, UWB) technology.· Define the user model, the interested hotspots model and the rules for detecting abnormal behaviors of the elderly. To get interested hotspots with a method based on DBSCAN algorithm. We proposed a method of statistics to get the two parameters which are set in advance. Building an alert line which is used to judge whether the duration that the elderly stayed in the monitored area can exceed the alert threshold.· To get activity patterns from data sequences and get the core activities which allow us to build the spatiotemporal model of the elderly's daily lives. We propose a method for abstractly expressing activity patterns.To sum up, in this paper we can get the elderly's trajectory data with the real-time positioning system which is mainly based on UWB technology, and it can judge whether the duration which the elderly stay in the monitoring area can exceed the threshold. The method for abstractly expressing the elderly's activity patterns could be used to understand unusual or abnormal activity.