Elderly-falling detection using distributed direction-sensitive pyroelectric infrared sensor arrays

Abstract

This paper presents a new distributed direction-sensitive infrared sensing approach for fall detection in elderly healthcare applications. Pyroelectric infrared (PIR) sensors are employed in sensing human activities. For capturing the characteristics of human normal and abnormal activities, three modules of a direction-sensitive PIR sensor are organized using a distributed sensing structure. The advantage of using the distributed sensing paradigm is that the synergistic motion patterns of head, upper-limb and lower-limb can be efficiently encoded and thus the more discriminative features can be captured. This is the new consideration of using PIR sensors in building a full detection system. In addition, a two-layer hidden Markov model is developed for recognizing a fall event based on the multidimensional signals of the distributed infrared sensing system. Experimental studies are conducted to validate the proposed method.