Font Size: a A A

Research On Information Extraction Method Of Speckle Interferometry Fringe Pattern

Posted on:2020-07-16Degree:MasterType:Thesis
Country:ChinaCandidate:Y N XueFull Text:PDF
GTID:2370330575459407Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
ESPI(Electronic Speckle Pattern Interferometry)technology is the main method to obtain speckle interferometry fringe pattern,which is suitable for vibration analysis,deformation and displacement measurement,non-destructive testing,strain stress detection of optically rough surface and tire inspection.It is a minimally destructive optical measurement technology.High-precision extraction of Speckle interference fringe pattern information is the key to ESPI application.We need to analyze and operate the Speckle interference fringe pattern carefully and parse the fringe into the physical quantity information of each point in the measurement field,mainly the displacement and phase shift.Then the information needed for detection or measurement of the relative application can be obtained.The essence of fringe pattern information extraction is to extract the quantitative distribution of the corresponding phase field in its gray distribution.Due to the special properties of speckle interferometry fringe pattern,its analysis and processing are different from other types of images.Up to now,some achievements have been obtained from the initial manual analysis and recording to the current relatively automated processing.The extraction methods are mainly divided into two categories: one is the fringe center-line method based on fringe brightness analysis,and the other is the phase method based on phase analysis of time and space.In recent years,the deformation phase extraction method of an object based on the theory of optical flow field has also attracted attention,and has been widely used because it does not require unwrapping and is suitable for small deformation detection.This thesis mainly studies the extraction of phase information of speckle interferometry fringe pattern,which is related to the fringe center-line method based on fringe brightness analysis,the phase method based on time and space phase analysis and the deformation phase extraction method based on optical flow field theory.Based on the in-depth comparative study of existing method and theory,a new method for deformation phase extraction is proposed,which can improve the accuracy of phase information extraction and is conducive to accurate interpretation of images.The main contents and innovations of the thesis are as follows:1.Image refinement is the most critical step in the fringe center-line method,in this thesis,the classical ESPI image refinement algorithm is improved.Phase epitaxy is introduced before the ESPI image is binarized,and the image after binarization is subjected to multiple nonlinear smoothing.The improved refinement algorithm adopts the serial processing operation mode.Although the speed is slow,it effectively improves the accuracy of phase information extraction.The burr and bifurcation of the image after refinement are significantly reduced,which is beneficial to the accurate interpretation of the fringe pattern information and provides great convenience for subsequent work and applications.2.The classical algorithms(phase shift method and Fourier transform method)for phase method are studied and compared.At the same time,the general phase shift algorithm GPSA(General Phase Shifting Algorithm)and the phase shift algorithm based on principle component analysis involved in the random phase shift algorithm are studied and simulated,which realizes the high precision solution of the deformation information of measured object under the condition of unknown phase shift and greatly reduces the requirements of external environmental stability and the high performance of the instrument system during the experiment.3.Based on the comparative study,an improved phase information extraction method based on optical flow field theory is proposed.The optical flow field HS(Hom Schunck)algorithm is used to evaluate the optical flow vector field between two consecutive images,and the fringe frequency is extracted by the improved adaptive window Fourier transform.And then the vector field and fringe frequency are utilized to calculate and obtain the full field deformation phase distribution of the object.The motion vector field and the phase deformation distribution are related and the process is simple and convenient.And the optimal window size is automatically adjusted as the fringe frequency changes,which can overcome the shortcoming of the window size fixing.It does not require phase unwrapping calculation and the phase deformation information extracted is more accurate when the fringe frequency is high.Also it is noise-insensitive and robust.
Keywords/Search Tags:electronic speckle interferometry, refinement algorithm, unwrapping, random phase shift algorithm, optical flow field
PDF Full Text Request
Related items