| Shack-Hartmann wavefront sensing technology is a high-precision optical measurement technique that utilizes wavefront phase information.It is widely used in fields such as human eye wavefront aberration measurement,medical diagnosis,astronomy,mirror surface shape detection,laser processing and optical manufacturing.This technology consists of two parts:spot centroid detection algorithm and wavefront reconstruction algorithm.Both of them will affect the accuracy of wavefront recovery.When the centroid position obtained by the spot centroid detection algorithm is fixed,different reconstruction algorithms will result in different reconstruction errors.This paper aims to study the problem of poor performance of spot centroid detection algorithm and wavefront reconstruction algorithm under low signal-to-noise ratio environment.The specific content and research results are as follows:(1)Aiming at the problem of low spot centroid detection accuracy of traditional centroid method and its improved algorithms under low signal-to-noise ratio environment,a correction algorithm based on sub-aperture spot centroid weighting is proposed.The algorithm is used to correct the Hartmann spot centroid position: when the spot information of a certain sub-aperture is unreliable,the weighted value of the previous position sub-aperture information is used as its estimated value;when it is reliable,the current sub-aperture information is used as its estimated value.In order to determine whether the spot information of the subaperture is reliable,the concept of correction threshold is introduced,and the selection of its value needs to balance system robustness and sensitivity.Simulation results show that this algorithm has reliability and effectiveness.(2)Aiming at the problem of the initial value problem in research content 1 and missing information for some sub-apertures,an optimized centroid correction algorithm based on sub-aperture information enhancement was proposed,which used cross window and 3×3window to collect sub-aperture information.The correction method in research content 1 relied on the reference value of the previous position of the Hartmann spot array,but if the reference value of the first position was unreliable,subsequent corrections would have problems.At the same time,the information collection method in research content 1 had limited information collection and poor system robustness,and could not handle situations where some sub-aperture spot information was missing.To solve these problems,the optimized centroid correction algorithm increased the number of reference sub-apertures and improved the amount of information collected.Simulation results show that the optimized centroid correction algorithm has obvious improvement on initial value problem and is better than traditional slope zeroing method on spot information missing problem,which can provide prerequisite conditions for wavefront reconstruction.(3)Combining research content 1 and research content 2,the problem of spot centroid detection in low signal-to-noise ratio environment was successfully solved.For the wavefront reconstruction problem,a method combining modal method and regional method was adopted.First,using part of spot centroid position information,modal method based on Zernike polynomials was used to obtain the low-frequency part of wavefront phase map.Then,using remaining spot centroid position information,regional method based on Shouthwell model was used to obtain the high-frequency part of wavefront phase map.Finally,the two parts were superimposed.This method not only ensured high accuracy,but also could understand relative size of low-frequency components through Zernike polynomial coefficients.It has great advantages in human eye aberration measurement field. |
