Research On Mobile Robot Automatic Target Extraction And Tracking Technology Based On Mean-shift Algorithm

Target tracking is an important part of computer vision.Target tracking technology has also been one of the hot topics in the field of computer vision research.This paper uses intelligent security robots as the engineering background to study the application of target tracking technology in mobile robots.A method of automatic target extraction and tracking based on traditional Mean-shift is proposed.Experimental verification is performed on the platform to achieve target tracking and control of mobile robots.The research content of this article can not only be applied to intelligent security robots,but also has application value in the fields of defense military,aerospace and daily life.In terms of mobile robot platforms,this article independently builds mobile robot experimental platforms based on the actual application scenarios of mobile robots.This paper has carried out kinematic analysis and structural design of mobile robots through the chassis and the gimbal,respectively.At the same time,the control method of the mobile robot system is researched in this paper.The transfer function is obtained by establishing the mathematical model of the mobile robot system to achieve precise closed-loop control.In terms of visual control cases,in order to solve the problem that traditional Mean-shift tracking algorithm needs to manually intercept the target before tracking,this paper proposes an automatic target extraction and tracking method based on the traditional Mean-shift tracking algorithm.This method integrates the Sum of Absolute Differences(SAD)matching method.It provides prior information of the target image for subsequent tracking,by identifying and matching the target object to be tracked in advance,then the target extraction and tracking can be achieved.A comparison experiment between the proposed method and the traditional Mean-shift tracking algorithm can verify that the proposed method can effectively achieve automatic target extraction and tracking.Finally,the method of automatic target extraction and tracking proposed in this paper is tested on the established mobile robot experiment platform,and the geometric coordinate data and depth information of the tracked target are obtained,which lays the foundation for the subsequent mobile robot control.