Statistical Analysis of Human Activity in Daily Life through a Data Set from the Infra-red Ray Sensors

Abstract

A statistical analysis is performed on the human activities in daily life . The data were obtained from seven infra-red ray sensors(attached by Ohta et al.)to several rooms at the single-living elderly home. It is shown that the position and movement of the person are finely reproduced, if a suitable pre-operation is performed on those data. A system for finding the abnormalities in activities is proposed based on the staying time distribution to characteristic states(places). Basic problems including the stationarity and the time-dependence of the statistics are investigated toward the realization of this system.