Research On The Measurement Method Of Micro-table Position And Pose Based On Binocular Vision

It is popular to use the method of non-contact measurement to detect the geometric quantities of MEMS at home and abroad since the structure and the loading abilities of MEMS are all very small, can not install sensors to measure the geometrical quantity. However the current measurement method is mainly used for the static measurement of three-dimensional shape and size, profile measurement and dynamic performance measurement, information detected can not be accurately described the position and pose of MEMS (such as: micro-mechanical motion platform), so it will not be able to test and evaluate the motion accuracy. In response to these circumstances, this paper presents a method of measuring the micro-table position and pose based on binocular vision, and then can get the micro-table position and pose in three-dimensional space and measure the kinematics error of the micro-table at every sampling time.Research contents and innovations in this paper as follows:1. MEMS are abstracted as a rectangular platform with regular shape in this paper, which is micro-table. The four corners of the micro-table are used as feature points, based on binocular vision, combined with coordinate transformation principle, the solving of the 3D position and pose of the micro worktable is established based on the characteristics, as the linear equation, analytical solutions exist, so no need to solve the iterative calculation.2. The camera calibration is researched, compared with the linear camera calibration and nonlinear calibration methods, the parameters of the camera calibration are processed. The non-linear camera calibration is established based on the Zhang's plane template calibration.3. Building the experimental platform based on the three dimensional measuring machine, extracting the corner of feature points, combination of coordinate transformation, getting the four corners of the world coordinates, finally, doing the position and pose of the micro-table simulation experiment, verifying the feasibility of the scheme. Based on the experiment, this paper discusses the factors affecting the measurement accuracy and demonstrates the feasibility of the method by theory.In short, this paper make the use of binocular vision to research the micro-table position and pose, and get the method for solving the position and pose based on the characteristics of the corners, as well as provide a theoretical basis on the follow-on work for error detection and assessment methods.