Indoor Person Identification and Fall Detection through Non-intrusive Floor Seismic Sensing

Abstract

This paper presents a novel in-network person identification and fall detection system that uses floor seismic data produced by footsteps and fall downs as an only source for recognition. Compared with other existing methods, our approach is done in real-time, which means the system is able to identify a person almost immediately with only one or two footsteps. An adapted in-network localization method is proposed in which sensors collaborate among them to recognize the person walking, and most importantly, detect if the person falls down at any moment. We also introduce a voting system among sensor nodes to improve accuracy in person identification. Our system is innovative since it can be robust to identify fall downs from other possible events, like jumps, door close, objects fall down, etc. Such a smart system can also be connected to smart commercial devices (like Google Home or Amazon Alexa) for emergency notifications. Our approach represents an advance in smart technology for elder people who live alone. Evaluation of the system shows it is able to identify people with one or two steps in an average of 93.75% (higher accuracy than other methods that use more footsteps), and it detects fall downs with an acceptance rate of 95.14% (distinguishing from other possible events). The fall down localization error is smaller than 0.28 meters, which it is acceptable compared to the height of a person.