Measuring Method Research Of Deformation Displacement About Structure Specimen In The Shaking Table Test Under Monocular Vision

Shaking table test is an important means on understanding and studying the seismic performance of building structures and structures and their failure mechanism under the action of seismic waves. The goal of shaking table test is to get deflection,displacement, velocity and acceleration motion parameters of specific test points on target object. The structure deformation is one of the important data in the test.In this thesis, the machine vision technology is applied to the test specimen deformation displacement measurement of shaking table test. The movement process about the structure specimen fixed on the vibration table is filmed by a video camera.The deformation displacement will be get after the follow-up image analysis. The measurement method is a complement of sensor measurement, providing more data sources from another channel. The main work of this thesis can be summarized as follows:(1) The thesis has studied the domestic and foreign research situation and trend about the machine vision technology applied to the displacement measurement. Then the basic theory of the machine vision technology applied to the deformation displacement measurement in the shaking table test has been understood, mainly including the imaging principle of digital image and its form and characteristics, the coordinates involved in image measurement, the relevant algorithm of image pre-processing, target positioning and calibration algorithm.(2) The measuring method of deformation displacement in the shaking table test based on the machine vision technology is proposed, and it also includes the concrete research on the key technology of this method. The thesis has analyzed performance of the general artificial sign and Retro-reflective targets, and designed the artificial marks firstly. Then, for the target and background adhesion the method based on mathematical morphology algorithm is adopted. Experiments show that the proposed algorithm can make good effect. After artificial landmark and background has completely separated, artificial landmark recognition algorithm has studied.According to the characteristics of the specimen deformation process, the deformation algorithm about other target points is proposed under the assumption that two target points located at the bottom of the specimens has no deformation. Finally, it has analyzed the factors influencing the measurement precision in the measurement method.(3) In view of the measurement method proposed by this thesis, it has designed the corresponding experiment schemes and do the experiments for experimentalverification in the vibration table lab. Firstly, the fixed displacement test is done to verify the measuring precision of this method. Then it has done the sine wave excitation test and used the deformation displacement data to make specimen deformation simulation animation, compared with cameras video filmed in the process of test to verify the feasibility of this method.