基于双目视觉信息的运动物体实时跟踪与测距-信号与信息处理专业论文.docxVIP

优秀毕业论文 精品参考文献资料 ABSTRACTABSTRACT ABSTRACT ABSTRACT：Mobile robot is all essential embranchment of robotics．With the rapid development of robotics，the application of mobile robot is enlarged．As the most sufficient source of information，computer vision makes the robot complete some intelligent function like human beings．In this paper,new algorithms to track and measure a moving object in 3D space are investigated for mobile robots with a binocular camera system．As the moving object out of the field of view,one or two cameras ofthe pan—tilt binocular system are controlled to search the object，and compute its position in a world coordinate system．Due to the limitation of the camera motion， the base of the mobile robot is then controlled to track the lost object．Tracking and measuring of an object can be done in real time as the cameras moving，without recalibrating the two cameras．Background modeling and color based Camshifi are used to identify moving objects．Experimental are shown to demonstrate the effectiveness of the proposed methods．I summarize the main work in this paper as below： Firstly,based on the Zhang’S methods，we make some amelioration,proposing an algorithms of parameter estimation．When the rotation of the camera occurred，we need not to recalibrate the camera，also we Can get the extrinsic parameter of the camera．It is convenient for the application on the mobile robot and makes the robot precept the information of the environment． Secondly,we make a comprehensive summary of the segmentation and tracking methods，especially analysis the algorithms of background modeling and color based Chamshifi．The traditional methods required that the camera call not been moved a lot or just tracking a specifically object that had been indicated before．Combining the advantage of these algorithms，we proposed the methods of the dynamic background arbitrary object tracking． Thirdly,we proposed the control strategy of mechanical part of the robot and make the system come true