Applied Study On Statistic Processing Of Optical Interferometric Fringe Pattern And Convolution Profilometry And Photoelastic Stress Analysis

Photomechanics is a science of technique, which is the combination of crystal optics, lodern optics, laser technique, mathematical statistics and digital image processing technique, the data and information of photomechanics are often represented by optical interferometric ringe patterns. The first major contribution of this dissertation is study on fringe pattern utomatic processing.It is significant in techniques and engineering applications to reduce noise and recovery he phase of interferometric fringe pattern. An adaptive filtering method, based on maximum losterior estimation rule and statistical signal processing technique, is presented. The new nethod can efficiently reduce Gaussian noise and noise with uniform distribution, and rather ccurately perform phase recovery from a single fringe pattern. The results of simulated and xperimental fringe patterns are presented to illustrate the processing of the new method.Secondly, considering the phase of fringe pattern can be recovered by solution of a equence of linear systems of equations, and the corresponding matrices are quite large and parse, the optimum numerical methods are available to solve this kind of systems. Among hese method, One of the simplest is the deepest gradient descent algorithm, but it has one disadvantage that the solutions are oscillatory and slowly convergent. A friction term is ntroduced to avoid oscillatory solutions and raise convergence rate.The third major contribution of this dissertation is study on 3-D diffuse object surface optical automatic measurement. It is important to automatically measure the 3-D object shapes, ecause topographic information is needed in automated manufacturing, production quality control, medicine, and Engineering. A robust algorithm is introduced in this paper. The algorithm, basing on convolution filtering technique and the fringes' gray values max-min distribution, can automatically and accurately carry out space frequency spectra shifting without spectra analysis and spectrogram in the technique of phase method of projected grating for 3-D object shapes measurement. The new algorithm makes image processing much Caster and the technique of Object shapes measurement more automated. The phase error caused by filtering and the crossed-optical-axes geometry setup is also discussed.In the end of this dissertation, the applied study of photoelasticity on the model stress analysis under its own weight and exterral loads is given. In the case of strength design of arge-scaled engineering structures, it is often necessary to consider the effect of the weight of the structure. In order to obtain the model stress under its own weight and external loads, the gravitational loading experiment must be performed. However, it is too expensive in investment, complex in experimental techniques, difficult and dangerous in performance. In this paper to overcome these drawbacks, the similarity relations together with a new methodand its principle of experiment have been developed based on the model theory of structural similarity. According to this principle, the total stress ;an be obtained as a sum of the stresses due to weight and external loading, which based on only one model experiment without any gravitational loading. By means of the new method in this paper, the 3-D photoelasticity analysis results of a bridge have been obtained, which well agree with the results of finite element method analysis.