A Smart Walking Stick for Gait Analysis of Elderly and People With Disabilities

Human gait analysis is of considerable significance in the assessment of a user’s movement status, abnormality warning and rehabilitation training. In this study, a smart walking stick (SWS) system is developed based on a single 9-axis attitude sensor for gait analysis of the elderly, people with disabilities, and other relevant users. The attitude of the SWS is solved by constructing an error vector of gravitational acceleration as an input to the angular velocity PID algorithm correction. An extended Kalman filter algorithm is used to make an optimal state estimate of the Euler angle. On the basis of such means, a threshold-based approach is adopted to binarize the angle of the SWS and divide the gait cycle, so as to combine the acceleration of the user’s movement and obtain information about the user’s gait, including the number of steps, stride duration, step length and step speed. The experimental results reveal that the accuracy of the number of steps can reach 100%, and the average error of stride duration, step length and step speed are 0.01±0.07s, -0.72±5.22cm and 0.75±3.74cm/s, respectively, with an average accuracy of 96.43%, 95.20% and 94.45%, respectively.