Research On Phase Recovery In Speckle Imaging

In the traditional optical imaging process,most optical equipment cannot directly detect the phase distribution of the light field,only the intensity distribution of the light field can be collected,and the phase information of the light field can be obtained by using the phase recovery algorithm through the measured light intensity,In order to get more information about the light field.From the 1970s to the present,the focus of basic research in modern optical technology has included phase recovery.This paper makes a detailed analysis and research on two types of classic phase recovery algorithms.And the practical application of phase recovery in the speckle autocorrelation imaging process is studied.Speckle imaging technology belongs to the basic application of phase recovery,and it has far-reaching influence and certain guiding significance on the application of phase recovery.The research content of this article mainly includes:(1)Various classical iterative algorithms to solve phase recovery are analyzed and researched.Summarized and discussed the basic principles of the classic phase recovery algorithm—GS(Gerchberg-Saxton)method,Error Reduction(ER)algorithm,Hybrid Input Output(HIO)algorithm,and angular spectrum iteration algorithm.Several algorithms have been simulated,and the advantages and disadvantages of the algorithms have been compared.The simulation results show that the GS algorithm is simple and fast,but the initial phase is uncertain,and the phase recovery result is also unstable.Both the ER algorithm and the HIO algorithm are optimized on the basis of the GS algorithm,and the convergence speed and convergence accuracy have been improved.The angular spectrum iteration method can solve the problem of phase recovery with diffraction distance due to the setting of the transfer function.(2)Starting from the development process,basic principles and mathematical models of the Transport-of-Intensity Equation(TIE),this article focuses on the methods for solving the light intensity propagation equation,including the Green function method,Zernike Polynomial method,fast Fourier transform method,etc.Through the Fourier transform method,the TIE equation can be quickly solved to obtain the phase information of the light field.According to the analytical results,the results of different approximations of the light intensity can be obtained,which can be divided into simple TIE(Simple TIE,STIE)and pseudo Precise TIE(Pseudo-exact TIE,PTIE),a detailed analysis and summary of the different situations of these two approximations.It can be seen from the theoretical derivation that the STIE method is simple and fast,but the result is inaccurate due to the loss of light and the multiple rounding of the curl term.Compared with STIE algorithm,the calculation efficiency of PTIE algorithm is low,but the result is more accurate.Through simulation and experiment,the accuracy of STIE method and PTIE method is compared,and the result is consistent with theoretical deduction.The accuracy of PTIE method is slightly higher than STIE method.(3)This article innovatively uses the angle spectrum iteration algorithm based on the HIO algorithm to combine with the more accurate PTIE method.The experimental and simulation results show that the HIO-PTIE method has higher accuracy and faster iteration speed.Because the advantages of the iterative method and the TIE method are different,the two methods can be combined to a certain extent,that is,after the initial phase is calculated by the TIE method,the result is regarded as the initial phase calculated by the iterative method for iterative calculation.The TIE method does not require iteration,and the calculation speed is fast,but the recovered phase result is not very accurate.The iterative method is precisely because the uncertainty of the initial phase causes the calculation result to be unstable,so the TIE method just compensates for this defect.According to the characteristics of different phase recovery algorithms,this paper divides the entire coherent light field diffraction area,and analyzes the diffraction areas suitable for different phase recovery algorithms.The experimental results show that the TIE method is suitable for near-field diffraction and the HIO-PTIE fusion algorithm is suitable.For mid-field diffraction,the iterative method is suitable for far-field diffraction.(4)Based on the speckle autocorrelation imaging theory,the causes and properties of speckle are discussed,and the speckle pattern of any image is numerically simulated.According to the Wiener Sinchin theorem and the imaging principle of image autocorrelation,the autocorrelation of speckle images is numerically simulated.The phase recovery algorithm is applied to the speckle autocorrelation imaging process,and the original target image is reconstructed through the speckle image obtained by numerical simulation.