Format of Original
Institute of Electrical and Electronics Engineers (IEEE)
2015 IEEE 39th Annual Computer Software and Applications Conference (COMPSAC)
The aim of this paper is to present a multisensory system that studies abnormal walking patterns to prevent a fall. Due to the growing elderly population, scientific research on smartphone-based gait detection systems has recently become an imperative component in decreasing elderly injuries due to falls. To address the issue of smart gait detection, we propose a gait classification system using smartshoe sensor data in this paper. We used shoe-worn pressure sensors on the foot and validated algorithms to extract the gait parameters during walking trials in a lab environment. This smartshoe contains four pressure sensors with a Wi-Fi communication module to unobtrusively collect data. To the best of our knowledge, this is the first system which can automatically detect abnormalities in walking patterns. A unique signal classification approach is presented by recognizing the abnormality in a subject's gait, and modeling the dynamics of a system as they are captured in a reconstructed phase space. Based on our experiments, we have found an 89% walking-based classification accuracy to help prevent falls.