Research on 3D Space Target Following Method of Mobile Robot Based on Binocular Vision

Abstract

In order to solve the problem of the target following in 3D (three-dimensional) space, the KCF and SGBM fusion algorithm is proposed. In this method, the position information of the target in the camera image is obtained by KCF algorithm, the depth information of the target is calculated by SGBM algorithm, and the three-dimensional coordinate of the target in camera coordinate system is determined. The binocular vision system is set on the mobile robot, and the mobile robot achieves target following through velocity information and angular velocity information. Experiments in different indoor environments show that the algorithm has low hardware requirements, good real-time performance, and high precision. It is suitable for target tracking by a robot in three-dimensional space.