Comparison of EKF and UKF target tracking algorithms based on observation distance

Abstract

Target tracking is one of the important research directions in the field of computer vision, and it is widely used in many fields such as precision guidance, human-computer interaction, and robot navigation. In this paper, a target tracking model is established, and the EKF and UKF target tracking algorithms based on observation distance are deeply studied, and simulation experiments are carried out respectively. The results show that when the nonlinearity of the system is not strong, the accuracy of the two algorithms is similar. In more complex situations, the UKF tracking target accuracy is higher.