Aberration correction method for Fourier lamination microscopy

Abstract

The invention discloses an aberration correction method for Fourier lamination microscopy. Due to the existence of the lens, aberration inevitably exists in an image acquired through a Fourier lamination microscopic optical path. The invention provides an updating strategy combining a Zernike polynomial and a two-dimensional matrix pupil function. Solving a Zernike polynomial coefficient aj while recovering the complex amplitude of the sample by the alternating projection method; constructing a pupil function according to the coefficient aj; then weighting and updating the two-dimensional matrix of the pupil function again according to the magnitude of the frequency spectrum energy; after iteration is completed, system aberration can be corrected, and a high-resolution complex amplitude image can be obtained. The method does not need hardware to correct aberration, and still has good robustness under the condition of serious aberration. And finally, the Fourier lamination microscopy improved by combining a Zernike polynomial and a two-dimensional matrix pupil function updating strategy is higher in aberration correction speed, larger in correction range and higher in precision.

Description

(1) Technical field

[0001] The invention belongs to the field of optical computational microscopy imaging and relates to an aberration correction method for Fourier stack microscopy. (2) Background technology

[0002] Throughout history, humans have been fascinated by the microscopic world. Optical microscopes are essential tools for exploring this microscopic world, playing an indispensable role in the advancement of human technology and civilization. With the rapid development of digital pathology, the demand for microscopes that combine high resolution with large field of view is increasing.

[0003] Fourier stacking microscopy came into being, and its appearance aroused everyone's expectations for a new generation of high-resolution, large-field microscopes with quantitative phase acquisition. Fourier stacking microscopy combines technical ideas such as phase recovery and synthetic aperture to solve the contradiction between the spatial resolution and imaging field of view of traditional microscopes. It can reconstruct high-resolution complex amplitude images while maintaining the original field of view of the objective lens. However, since the lenses of the microscope system will introduce aberrations, this affects the improvement of resolution and the size of the effective field of view. The physical method of aberration correction is hardware correction, but hardware correction (such as deformable mirrors) is expensive and slow to respond, and it also increases the complexity of the experimental work.

[0004] Compared to the high cost of hardware correction, computer software-based aberration correction reduces costs and improves efficiency, effectively estimating and correcting complex aberrations. This method not only improves the quality of reconstructed images but also offers greater robustness and is applicable to a wide range of aberration types. Depending on actual needs, GPU-based parameter optimization can be used to increase speed by more than 10 times. (3) Summary of the invention

[0005] The present invention proposes an aberration correction method for Fourier stack microscopy, aiming to solve the problem of system aberration and improve the correctable range of aberration.

[0006] To achieve the above object, the present invention provides a method for correcting aberrations in Fourier stack microscopy, the method comprising the following steps:

[0007] S1: Using a Fourier stacking microscope system, low-resolution images of the sample are collected by LED illumination at different angles, where the sample image is affected by system aberrations.

[0008] S2: Use the alternating projection method to update the low-resolution image and pupil function of the sample and update the sample spectrum;

[0009] S3: Calculate the coefficient a of the Zernike polynomial based on the difference between the spectrum and image before and after the update j , through the polynomial coefficient aj Constrained construction of pupil function;

[0010] S4: In the polynomial coefficient a j Based on the two-dimensional matrix of the modeled pupil, a weighted value is assigned according to the energy of each pixel in the S2 updated spectrum, and the two-dimensional matrix of the pupil is updated again;

[0011] S5: Repeat steps S2-S4, update the sample complex amplitude and pupil function during the alternating projection algorithm, and obtain a high-resolution image after reaching the number of iterations.

[0012] In S1, the Fourier stacking microscope system consists of an LED array light source, a microscope objective lens, a tube lens, and a CMOS camera. The captured image has system aberrations such as astigmatism, defocus, and spherical aberration.

[0013] In S2, the sub-region spectrum is updated using the alternating projection method combined with the pupil function, wherein the ideal pupil model during the first update is expressed as:

[0014]

[0015] The model can be regarded as a low-pass filter, where NA represents the numerical aperture and λ represents the wavelength. It can be understood that only light within the numerical aperture range can be received by the optical system. It is the coordinate transformation of the pupil function P(x,y) in the frequency domain. (k x ,k y ) represents the two-dimensional frequency domain coordinate system.

[0016] Furthermore, the pupil function with aberration is modeled as:

[0017]

[0018] Among them, Z j (ρ,θ) are Zernike polynomials used to describe different types of aberrations, a j is the corresponding coefficient. We assume that P(u) is a pure phase object, and the influence can be eliminated by multiplying it by its conjugate during the spectrum update process.

[0019] In S3, Zernike polynomial coefficients are calculated based on the difference between the image obtained by inverse spectrum transformation updated in S2 and the original image, and the coefficients are used to model the pupil function.

[0020] Furthermore, the pupil function modeled by the coefficients of S3 is used to assign a weighted value according to the energy of each pixel point in the updated spectrum, and the pupil function is regulated by the weighted value to be updated again.

[0021] Furthermore, steps S2-S4 are repeated to update the sample complex amplitude and pupil function in the alternating projection algorithm. After the iteration is completed, the restored pupil function and high-resolution complex amplitude image are obtained.

[0022] The beneficial effects of the present invention are:

[0023] The present invention designs a method for aberration correction in Fourier stack microscopy. The method uses an update strategy that combines Zernike polynomials and two-dimensional matrix pupil functions, which greatly improves the efficiency of pupil recovery compared to traditional methods. The method solves the Zernike polynomial coefficient a during the implementation of the alternating projection method. j ; According to the coefficient a j The pupil function is constructed; then, the two-dimensional matrix of the pupil function is updated again based on the weighted spectral energy to obtain the pupil function. This method combines Zernike polynomials with a two-dimensional matrix pupil function update strategy, resulting in faster aberration correction, a wider correction range, and higher accuracy. It also maintains good robustness even in the presence of severe aberrations. (IV) Description of the accompanying drawings

[0024] Figure 1 1. It is a schematic diagram of an algorithm flow of an aberration correction method for Fourier stack microscopy according to the present invention;

[0025] Figure 2 Schematic diagram comparing the aberration correction results of the present invention; wherein, (a1) is a simulated low-resolution image affected by aberration, (a2) is a high-resolution image restored by the present invention, and (a3) is a corrected pupil with large aberration.

[0026] Figure 3 It is a schematic diagram comparing the results of the first five iterations of the present invention; wherein, (b1) is the image restoration result and the Zernike polynomial coefficient fitting result of the alternating projection Zernike polynomial modeling pupil restoration method, and (b2) is the image restoration result and the Zernike polynomial coefficient fitting result of the method of the present invention. (V) Specific implementation methods

[0027] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be understood as limiting the present invention.

[0028] See also Figure 1 The present invention provides a method for correcting aberrations in a Fourier stack microscope, comprising the following steps:

[0029] S1: Using a Fourier stacking microscope system, low-resolution images of the sample are collected through LED illumination at different angles. The sample image is affected by system aberrations.

[0030] S2: Use the alternating projection method to update the low-resolution image and pupil of the sample and update the sample spectrum;

[0031] S3: Calculate the coefficient a of the Zernike polynomial based on the difference between the spectrum and image before and after the update j , through the polynomial coefficient a j Constrained construction of pupil function;

[0032] S4: In the polynomial coefficient a j Based on the two-dimensional matrix of the modeled pupil, a weighted value is assigned according to the energy of each pixel in the S2 updated spectrum, and the two-dimensional matrix of the pupil is updated again;

[0033] S5: Repeat steps S2-S4, update the sample complex amplitude and pupil function during the alternating projection algorithm, and obtain a high-resolution image after reaching the number of iterations.

[0034] The following is a further explanation based on the specific implementation steps:

[0035] In step S1, the Fourier stacking microscope system consists of an LED array light source, a microscope objective lens, a tube lens, and a CMOS camera. During the acquisition process, the LED array light source is lit in sequence according to the update order. After the light waves of the LEDs at each position pass through the sample, the microscope objective lens, and the tube lens, their intensity information is recorded on the CMOS camera. When the set LEDs are lit and recorded in sequence, the acquisition is completed.

[0036] In step S2, the alternating projection method is combined with the pupil function to update the sub-region spectrum, where the ideal pupil model during the first update is expressed as:

[0037]

[0038] The model can be regarded as a low-pass filter, where NA represents the numerical aperture and λ represents the wavelength. It can be understood that only light within the numerical aperture range can be received by the optical system. It is the coordinate transformation of the pupil function P(x,y) in the frequency domain. (k x ,k y ) represents the two-dimensional frequency domain coordinate system.

[0039] The pupil function with aberration is modeled as:

[0040]

[0041] Among them, Z j(ρ,θ) are Zernike polynomials used to describe different types of aberrations, a j is the corresponding coefficient. We assume that P(u) is a pure phase object, and the influence can be eliminated by multiplying it by its conjugate during the spectrum update process.

[0042] In step S3, the Zernike polynomial coefficients are calculated based on the difference between the image obtained by the inverse spectrum transformation updated in step S2 and the original image, and the coefficients are used to model the pupil function.

[0043] In step S4, after obtaining the two-dimensional matrix of the pupil function in S3, a weighted update is performed based on the energy of the pixels in the updated spectrum of S2. Different pixels are assigned different weights, and the pupil is updated again to effectively improve the recovery of high-frequency components.

[0044] In step S5, steps S2-S4 are repeated, and the sample complex amplitude and pupil function are updated in the alternating projection algorithm process. After the number of iterations is reached, a high-resolution image is obtained.

[0045] The above disclosure is only a preferred embodiment of the present invention, and certainly cannot be used to limit the scope of the rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made in accordance with the claims of the present invention are still within the scope of the invention.

Claims

1. A method for correcting aberrations in Fourier stack microscopy, characterized in that: The method comprises the following steps: S1: Using a Fourier stacking microscope system, low-resolution images of the sample are collected by LED illumination at different angles, where the sample image is affected by system aberrations; S2: Use the alternating projection method to update the low-resolution image and pupil function of the sample and update the sample spectrum; S3: Calculate the coefficient a of the Zernike polynomial based on the difference between the spectrum and image before and after the update j , through the polynomial coefficient a j Constrained construction of pupil function; S4: In the polynomial coefficient a j Based on the two-dimensional matrix of the modeled pupil, a weighted value is assigned according to the energy of each pixel in the S2 updated spectrum, and the two-dimensional matrix of the pupil is updated again; S5: Repeat steps S2-S4, update the sample complex amplitude and pupil function during the alternating projection algorithm, and obtain a high-resolution image after reaching the number of iterations.

2. The aberration correction method for Fourier stack microscopy according to claim 1, characterized in that: The aberration correction method for Fourier stack microscopy uses a strategy combining Zernike polynomials and two-dimensional matrix pupil function update to correct system aberrations while restoring the sample complex amplitude.

3. The aberration correction method for Fourier stack microscopy according to claim 1, characterized in that: In S1, the Fourier stacked microscope system is composed of an array LED light source, a microscope objective lens, a tube lens, and a CMOS camera.

4. The aberration correction method for Fourier stack microscopy according to claim 1, wherein: In S2, the process of alternating projection includes initial guessing, calculating the low-resolution light field corresponding to the imaging plane, updating the amplitude of the light field, updating the spectrum of the object function, and updating the pupil function.

5. The aberration correction method for Fourier stack microscopy according to claim 1, characterized in that: In S3, the system pupil with aberration is modeled by Zernike polynomials as follows: Among them, (k x ,k y ) represents the two-dimensional frequency domain coordinate system, Z j (ρ,θ) are Zernike polynomials used to describe different types of aberrations, a j is its corresponding coefficient. We assume that P(u) is a pure phase object, and the aberration effect can be eliminated by multiplying it by its conjugate during the spectrum update process.

6. The aberration correction method for Fourier stack microscopy according to claim 1, characterized in that: In S3, the pupil function is represented by the calculated Zernike polynomial and its coefficients.

7. The aberration correction method for Fourier stack microscopy according to claim 1, characterized in that: In S4, the two-dimensional matrix of the pupil function is represented by the Zernike polynomial coefficients obtained in S3 and the spectrum update value obtained in S2, and weighted update is performed again according to the magnitude of the spectrum energy.

Images