| Summary: | 碩士 === 中華大學 === 機械與航太工程研究所 === 93 === Recently, 3D computer vision has been widely used in the application areas like robot vision, industrial metrology, and measurement of deformations in material. Hence camera calibration techniques were developed to acquire the relationship between the 3D object coordinates and the image coordinates. In the paper a new method of camera calibration is proposed to calibrate a stereo vision to measure the deformation field of a material with micron measurement accuracy.
The center positions of circle dots were used as calibration points. Since the projection of the circle dot on the image plane is an ellipse, the relationship between the center of the circle dot and the center of the ellipse was obtained mathematically in the paper to reduce the bias between the camera model and the observed center of the ellipses. In addition, both of the radial and decentering lens distortions are included in the camera model equation to improve the accuracy of the model equation.
After camera calibration, the tensile tests on the double dots specimen, the dot grids specimen, and the random pattern specimen were conducted and the calibrated stereo vision was used to measure the deformation field of the specimens and calculate the axial strain, lateral strain, and Poisson’s ratio to evaluate the ability of the stereo vision to measure a deformed body.
|
|---|
