Patterns of Bipedal Walking on Tri-axial Acceleration Signals and Their Use in Identifying Falling Risk of Older People

Falls are the leading cause of injury-related death and hospitalization of aging people. To clarify this issue, temporal-spatial differences of gait between young people and the aging people with high falling risks are well investigated by many authors. Body acceleration is one of the measures receiving most interests, but major relevant literatures studied it only on one axis - mainly vertical direction. In this paper, we study the triaxial acceleration signals during body moving and find some patterns inside. Based on the patterns, we propose a method to identify walking waveforms within the signal of body acceleration, and derive some measurements for assessing the rhythmicity in walking. The experiments show that our measurements are effective in identifying aging people with walking problems.