UKF based moving target localizing and tracking method in 3D environments

Abstract

Simultaneously localizing and tracking of moving target is one of the most important applications of mobile robot system. However, single robot often presents great limitation when facing such tasks because the equipped sensors are often with low precision considering factors of loading ability and cost. Thus, in this paper, cooperative strategy of multiple robot systems is researched to handle this problem. Firstly, 3D localizing algorithm of two robot systems using vision sensors is proposed based on the basic theory of stereo vision, which is then modeled as a nonlinear estimation problem and solved by using UKF algorithm. Subsequently, formation control algorithm is introduced to make multiple robots track to the moving target. Finally, experiments are conducted and the results are analyzed with respect to an indoor multiple-flying-robot platform to verify the feasibility and validity of the proposed algorithms.