Recognization of Stance Phase Using Flexible Pressure Sensors

Abstract

Walking is a periodic activity in daily life. A mount of lower limb diseases can be reflected in abnormal gait. Recognization of the gait phase, especially the stance phase, benefits not only the clinical rehabilitation but also some human-machine system control such as exoskeleton robots. The study deals mainly with the recognization of stance phase by using a pair of in-shoe flexible pressure sensors that collected the real time plantar forces and a camera-based motion system to get the accurate stance phase. Furthermore, a Back-Propagation neural network is utilized in order to get the relationship between the foot force and the gait phase. The result shows that the sub-stance phases can be recognized by foot force with an accurate rate of 92.07%, which proves that the plantar force is closely related to gait phase, especially the stance phase, and the sensor and the method proposed can be applied in clinical rehabilitation and some human-machine system with high reliability.