| Wavenumber Scanning Interferometry（WSI）is a high-precision tomographic measurement technology,which combines wavenumber scanning and full-field measurement.And by using the phase contrast,WSI is able to perform tomographic measurement of the deformation field and strain field inside the material.Currently,it is a photo-mechanics technology with great development potential and application prospects in modern optical detection.Although WSI has shown great performance in the tomographic measurement of the deformation field and strain field inside the composite material,there are still two problems that need to be solved in the application:（1）The narrow range of the laser’s wavenumber with non-mode-hopping results in insufficient depth measurement resolution of the system,and it cannot effectively evaluate the interference information inside the measured material;（2）Affected by speckle decorrelation,the SNR of WSI strain reconstruction is low in the case of large deformation.This article first analyzes and discusses the relevant principles of WSI,and then introduces the WSI measurement system built by author’s laboratory.In order to solve the above problems,this article will conducts theoretical research and method verification from the following two aspects:（1）This article attempts to perform wavenumber synthesis on multiple non-modehopping wavenumber scanning sections of the laser,and proposes to use Real-valued Iterative Adaptive Approach（RIAA）to demodulate interference data.This method constructs a sparse optimization model of the interference spectrum and uses the interference spectrum matrix optimization to evaluating the amplitude-frequency and phase-frequency information of the interference signal.This method can effectively suppress abnormal sidelobes,and greatly improve the tomographic resolution of the system without changing the WSI hardware system.The experimental results show that the RIAA can successfully demodulate the internal information of the interference signal and improve the measurement accuracy of the system.Moreover,it can reach or even exceed the depth measurement resolution of the standard transform domain method under the same bandwidth of laser continuous scanning as synthetic wavenumber sequence.（2）An adaptive incremental method for the cumulative strain estimation in phase data is proposed.The method firstly counts the amount of phase noise points by mapping a binary noise map.After the noise threshold value is preset,this method adaptively adjusts the interframe interval in terms of the phase-to-noise ratio,and then uses the vector method to calculate the strain of the phase difference between two frames and accumulates multiple strain fields during the deformation process.Finally,while reducing the cumulative number,the efficient estimation of cumulative strain is implemented,which overcomes the problem of the low strain signal-to-noise ratio（SNR）caused by speckle decorrelation.Taking advantage of the fact that the level of phase noise is related to the different strain rates in accordance with the speckle decorrelation,the proposed method can estimate the large strains with high computation efficiency as well as SNR enhancement in nonlinear change of sample deformations.Real experiments of polymer curing were performed to prove the high robustness of adaptive incremental method in estimating the large strains.The proposed method expands the practicability of the incremental method in more complex scenes. |
