Study On The Method Of 3D Digital Image Correlation In Experimental Photomechanics

Three-dimensional measurement for the shape and deformation of the surface has been widely applied in the fields of aerospace and machine manufacture. With the increasing requirements of measurement precision and computation efficiency from various industries, there are some key technologies that need to be solved in 3D digital image correlation method based on stereo vision. The study in this thesis aims to improve the measurement performance of 3D digital image correlation method. The main contents and achievements are as follows:The effect of camera shoot attitude on the measurement accuracy is studied first, including the error analysis in the case of slant optical axis, and the analysis of the relationship between the measurement error with the base line and the camera angle in stereo vision system. These will provid guidance on the high-precision design of camera attitude. The method of corner detection in 3D digital image correlation method for different types of calibration boards is studied and improved. Programs on these methods which contribute to analyse the calibration error and the precision of corner detection have been realizated. The system parameters of stereo camera are readily obtained with the result of corner detection. The simulation method in spectral domain for speckle pattern translation and deformation is proposed in this paper, with the study on the evaluation indicator of the speckle quality, providing the guidance for simulation and evaluation of the speckle pattern in engineering. The 2DCWT based correlation in spectral domain is studied and aims to improve speed, accuracy and anti-noise property of identification. Finally the 3-D coordinates and the equations of the calibration board were obtained, and the surface profile of the bending plate was built through 3-D reconstruction. All of these ensure the realization of the ultimate goal of the 3D digital image correlation method.All the investigations above provide effective measuring methods in 3-D digital image correlation with higher measurement precision and computing speed. These achievements are of great value in experimental research and engineering application fields.