Efficient jitter compensation using double exponential smoothing

This paper proposes a new jitter reduction scheme based on double exponential smoothing (DES). We compare this DES-based method to jitter reduction methods based on the Kalman filter (KF) and extended Kalman filter (EKF), two well-known methods of jitter reduction. To evaluate the jitter reduction performance, we used a laser pointer interaction system with a known problem with jittery laser spot movements caused by natural hand tremors. We show that the DES-based scheme runs approximately 100 times faster than the EKF-based method and 19 times faster than the KF-based method. Furthermore, in terms of jitter reduction, the proposed DES-based method yields approximately 18% better results than the EKF-based method and 20% better results than the KF-based method.