Study On Key Techniques Of Precision Portable Vision Metrology

Facing the needs for precision measurement in advanced manufacturing industry, the key techniques of precision portable vision metrology are studied to achieve static measurement of three-dimensional coordinates in large equipment manufacturing. In portable vision metrology, single hand-held camera is used to take images of measured objects from various positions and directions. Then bundle adjustment optimization model is constructed based on the ray intersection constraint and solved by non-linear mathematical method to get 3D coordinates. Relevant principle and implementation method is studied, and then corresponding hardware system is set up. On this basis, emphasis is put on the key techniques such as how to improve measuring accuracy and expand the application. It provides theoretical and technological basis for improving measuring accuracy and developing new methods to satisfy the specific application requirement. The main work of this thesis is as following:1. The needs and characters of precision measurement in advanced manufacturing are summarized. The generation and development history of portable vision metrology is discussed. A large number of relevant applications are enumerated. Then the origin and significance of the subject is explained based on application needs, technological development and research foundations. Finally, the implementation and key techniques of portable vision metrology is described.2. It studied the principle and implementation of portable vision metrology thoroughly and put emphasis on its key techniques which include the constructing and solving of ray bundle adjustment model based on imaging model of camera, fitting image processing method, precision extracting of centroid and identifying of code targets, automatic orientating of measuring station based on code targets, matching homonymy points based on epipolar geometry and improved matching method. Relevant measuring system is set up and experiment has been achieved to verify the technical feasibility and accuracy.3. The optimization accuracy of ray bundle adjustment is studied. Considering the geometric relation of ray intersection constraint, error factors of ray bundle adjustment are separated in real space. The effect of imaging model's error on optimization accuracy is analyzed deeply by means of simulation and experiment. And the effect of difference between ray bundle adjustment models on optimization accuracy is studied which includes whether the internal parameters of camera are diverse and different intersection constraints of various measuring stations. Verification experiment is achieved and conclusions are summarized to provide basis for improving the accuracy of moving vision metrology.4. An improved portable vision method which refrains from using code targets is studied to meet specific application requirements. After expatiating the whole implementing process, automatic orientation of measuring station is studied intensively. Then the error and feasibility is analyzed and verified by experiments.5. The portable vision method for industry robot is studied to comform to the features of flexible manufacturing. The automatic orientation of measuring station based on the kinematics model of robot is studied and relevant experiment is achieved to verify its feasibility. Considering the positioning error of robot, a correcting method based on 5 points perspective location (P5P for short) algorithm is studied. Because the accuracy of P5P algorithm is low, the error characteristics of robot is analyzed and corresponding improving method is studied. Finally, relevant simulation experiment is achieved to verify its feasibility and accuracy.