Font Size: a A A

Study On Motion Simulator Pose Measurement System Based On Computer Vision

Posted on:2013-06-03Degree:MasterType:Thesis
Country:ChinaCandidate:Q ZhangFull Text:PDF
GTID:2248330377458419Subject:Measuring and Testing Technology and Instruments
Abstract/Summary:PDF Full Text Request
This paper is arranged under the application of spacecraft rendezvous and dockingtechnologies. Our purpose is to solve the real-time measurement of relative position andrelative attitude between objective and tracking devices, when they are nearby. In thedissertation, firstly the ptical measurement principium was elaborated based on binocularstereo vision system. And then the feasibility was proved of this method in theory. Secondly,the theoretical derivation of relative position and relative attitude will be presented.The technology of the space docking mechanism was always studied by researchersaround the world. With the development of the computer vision theory the modernmeasurement techniques, especially the non-contact measurement technology has a rapiddevelopment. Computer vision makes the computer recognize the three-dimensionalenvironment by the two-dimensional images, including its shape, Position, Posture,movement and so on. More over it can carry through the description, memory, recognitionand understanding to them. The research target of binocular stereo vision is that makesmachine have the ability that can recognize three dimensions environment information fromtwo dimensions image information, which can not only perceive geometric information in thethree dimensions environment, but also can process this information, including description,storages, recognition and comprehension.Firstly, a system of measurement is proposed and implement in this paper, which isbased on a theory of binocular stereo vision. The author establish system model, and optimizethe parameters and methods of system distribution station. The knowledge about feature pointextracting and stereo matching was described in the paper.Secondly, in software part, advanced my own idea to fetch and match feature points on theimage. Convolution is introduced in the calculation of normalized correlation coefficient toenhance the speed. Finally the paper developed a software test system making use of Matlab.Finally, the result shows the most error of target’s liner move; the error of attitude angle isbelow. The method in this paper has many advantages, including effective, practical use,simple quick, real-time effective and high precision and so on which provide guarantee forthe late work.
Keywords/Search Tags:computer vision, docking mechanism system, pose measure, scale invariantfeature transform (SIFT), stereo matching, feature point extraction
PDF Full Text Request
Related items