Object Tracking Algorithm Based on Better Parallax of Monocular under Unknown Environment

Abstract

Under unknown environment, the problem of simultaneous localization, mapping and object tracking based on monocular vision needs sufficient parallax to accurately estimate the state of the target. Therefore, this paper proposes an object tracking algorithm based on better parallax. In order to meet the parallax requirement, the algorithm utilizes the direction in which the target ellipsoidal projection area changes the most as the direction of better parallax. In addition, considering the energy efficiency, the speed of the robot should be reduced in engineering applications. For the different parallax speeds, some simulation experiments are carried out to testify the proposed algorithm. The results show that compared with the heuristic tracking algorithm, the proposed algorithm has better performance and it can also satisfy the parallax requirement of the object tracking with a slow speed, which demonstrates the high project application value of the proposed algorithm.