A novel estimation technique based on curve fitting and sigma-point Kalman filtering with time-delayed measurements

摘要

In this paper, we present a novel mobile target tracking filter that incorporates delayed measurements captured by a video camera onboard a flying platform, such as Unmanned Aerial Vehicles (UAVs). A curve-fitting-based technique is developed to characterize the system state information in the past. The coefficients are obtained and updated using an Extended Kalman Filter (EKF) based technique, which requires a much smaller amount of storage memory compared to a buffer based method. A filtering technique is then developed based on the EKF-fitted curve and the Sigma Point Kalman Filter (SPKF) technique. The simulations results show that the proposed tracking filter outperforms two other existing tracking filters, out-of-order SPKF and augmented-state SPKF.