VAMPIR- an automatic fall detection system using a vertical PIR sensor array

Abstract

Falling is a common health problem for elderly. It is reported that about 12 million adults 65 and older fall each year in the United States. To address this problem, at the Center for Eldercare and Rehabilitation Technologies in the University of Missouri we are investigating multiple fall detection systems. In this paper, we present an automatic fall detection system called VAMPIR based on a vertical array of multiple passive infrared (PIR) sensors. PIR sensors provide an inexpensive way to recognize human activity based on its infrared signature. To differentiate between falls and other human activities such as walking, sitting on a chair, bending over etc., we used a pattern recognition algorithm based on hidden Markov models (HMM). We obtained encouraging classification results on a pilot dataset that contained 42 falls and multiple non-fall human activities performed by trained stunt actors.