Modeling human gait using a Kalman filter to measure walking distance

Abstract

In this demo, we present a novel method to estimate joint angles and distance traveled by a human while walking. Understanding the kinematics of the human leg gives the velocities associated with forward human motion. Gyroscopes and accelerometers placed at two limbs provide the required measurement inputs. The inputs are used to estimate the desired state parameters associated with forward motion using a constrained Kalman Filter. Experimental results with walking subjects show that distance walked can be measured with accuracy comparable to state of the art motion tracking systems. The average RMSE is 0.05 meters per stride, which corresponds to 95% accuracy considering average stride length of 1 metre from the experiments.