Estimation of indoor physical activity level based on footstep vibration signal measured by mems accelerometer in smart home environments

Abstract

A smart home environment equipped with pervasive net- worked-sensors enables us to measure and analyze various vital signals related to personal health. For example, foot stepping, gait pattern, and posture can be used for assessing the level of activities and health state among the elderly and disabled people. In this paper, we sense and use footstep vibration signals measured by floor-mounted, MEMS accelerometers deployed tangent to wall sides, for estimating the level of indoor physical activity. With growing concern towards obesity in older adults and disabled people, this paper deals primarily with the estimation of energy expenditure in human body. It also supports the localization of footstep sources, extraction of statistical parameters on daily living pattern, and identification of pathological gait pattern. Unlike other sensors such as cameras or microphones, MEMS accelerometer sensor can measure many biomedical signatures without invoking personal privacy concerns. © 2009 Springer Berlin Heidelberg.