Coherent diffraction imaging automatic focusing method and system

Abstract

The invention belongs to the field of coherent diffraction imaging, and particularly discloses an automatic focusing method and system for coherent diffraction imaging. Multiple groups of to-be-detected samples at different positions are simulated through the near-field propagation model, wavelet transformation is performed on the simulated to-be-detected samples to obtain the wavelet coefficient, the definition is calculated according to the wavelet coefficient, and the axial distance is weighted and updated by using the definition of all the simulated to-be-detected samples so as to gradually approach the actual value of the axial distance between the sample and the detector. And automatic focusing of coherent diffraction imaging is realized. According to the method, the definition of the reconstructed image is calculated by adopting an operator in a frequency domain for the first time, so that the method has higher axial contrast and noise robustness, and the convergence speed, precision and robustness are greatly improved; through wavelet transform, detail components of the image in the horizontal direction, the vertical direction and the diagonal direction are extracted in the frequency domain so as to evaluate the definition of the reconstructed image, the weights in different directions can be flexibly adjusted, and the method has better generalization.

Description

Technical Field

[0001] The present application belongs to the technical field of coherent diffraction imaging, and more specifically, to a coherent diffraction imaging automatic focusing method and system. Background Art

[0002] Coherent diffraction imaging is a lensless phase imaging technology. The detector records the diffraction field after the illumination probe and the sample to be tested in the reciprocal space, and the complex amplitude information of the illumination probe and the sample to be tested is solved by an iterative algorithm to achieve diffraction-limited resolution. Since coherent diffraction imaging does not rely on lenses, it has broad application prospects in the imaging field of radiation sources such as extreme ultraviolet, X-ray and electron beam imaging, which are difficult to produce high-quality optical imaging elements. It has been widely studied and various improved forms have emerged.

[0003] The focal length of coherent diffraction imaging refers to the axial distance between the sample to be measured and the detector target surface, which will directly affect the clarity of coherent diffraction imaging and is the main source of error in coherent diffraction. The wrong axial distance makes the light field propagation distance calculated in the coherent diffraction imaging algorithm different from the actual distance, resulting in the light field information in the algorithm being unable to correspond to the real light field, and then making the coherent diffraction reconstructed image blurred due to scaling and artifacts; in coherent diffraction imaging using multiple diffraction field information, the wrong axial distance will further cause the relative relationship between the diffraction fields during calculation to be inconsistent with the actual relationship, further reducing the reconstruction quality of coherent diffraction imaging. However, the axial distance is difficult to measure and calibrate accurately, and how to obtain an accurate focal length is a major problem in coherent diffraction imaging.

[0004] The coherent diffraction imaging autofocus algorithm proposed in recent years, namely the coherent diffraction imaging autofocus algorithm based on sharpness (see Optics Letters, 2020, 45(7): 2030-2033) and the coherent diffraction imaging autofocus algorithm based on adaptive total variation (see Optics and Lasers in Engineering, 2022, 158, 107136), evaluates the clarity of the reconstructed image at multiple axial positions by calculating the 2-norm or p-norm (p is a value that changes adaptively according to the image intensity) of the image gradient in the spatial domain, and gradually searches for and approximates the true axial distance, thereby realizing axial distance calibration without the assistance of other additional equipment.

[0005] However, the image clarity evaluation method in this type of spatial domain has limited axial contrast, poor noise robustness, and few evaluation dimensions, which makes the existing diffraction imaging autofocus algorithm have defects in convergence speed, convergence accuracy, initial error robustness, and generalization. Summary of the invention

[0006] In view of the defects of the prior art, the purpose of the present application is to provide a coherent diffraction imaging automatic focusing method and system, which aims to solve the problem of axial distance error between the sample to be measured and the detector in the coherent diffraction imaging reconstruction process.

[0007] To achieve the above objectives, in a first aspect, the present application provides a coherent diffraction imaging automatic focusing method, comprising:

[0008] S1. Obtain the distance to be calibrated and the reconstructed sample to be tested and the illumination probe at this axial distance;

[0009] S2. Using the near-field propagation model, the reconstructed sample to be tested at the current axial distance is propagated for several distances to obtain simulated samples to be tested at different calibration positions;

[0010] S3. Perform wavelet transform on each simulated sample to be tested to obtain corresponding wavelet coefficients, wherein the wavelet coefficients include approximation coefficients, horizontal detail coefficients, vertical detail coefficients and diagonal detail coefficients;

[0011] S4. Calculate the variance of the wavelet coefficients of each simulated sample to be tested and sum them to obtain the corresponding clarity;

[0012] S5. Using the clarity of all simulated samples to be tested as weights, weighted update of the current axial distance, and update of the reconstructed sample to be tested and the illumination probe at the current axial distance;

[0013] S6. Repeat S2-S5 until the current axial distance approaches the actual value of the axial distance, and use the current axial distance as the calibrated axial distance to achieve automatic focusing of coherent diffraction imaging, while obtaining a clear reconstruction of the sample to be tested and the illumination probe after calibration.

[0014] Preferably, step S2 is specifically as follows:

[0015] S21. Obtain the current axial distance and calibration step, set multiple calibration positions along the axial direction with the current axial distance as the center according to the calibration step, and obtain the propagation distance of each calibration position relative to the center point;

[0016] S22. Use the near-field propagation model to propagate the sample to be tested reconstructed at the current axial distance to different distances to simulate the sample to be tested reconstructed by the coherent diffraction imaging algorithm at different calibration positions.

[0017] Preferably, the calibration step is an integer multiple of the depth of field of the coherent diffraction system.

[0018] Preferably, step S3 is specifically as follows:

[0019] [cA mm , cH mm , cV mm , cD m]=dwt2(O m )

[0020] Among them, cA mm , cH mm ,cV mm and cD mm denote the approximate coefficient matrix, horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image at the mth calibration position, dwt2 denotes a two-dimensional discrete wavelet transform function, O mm It indicates that the mth calibration position simulates the sample image to be tested, m=1, 2, ..., M, where M is the number of calibration positions.

[0021] Preferably, the calculation formula of the clarity is as follows:

[0022]

[0023] in, represents the clarity of the image of the sample to be tested at the mth calibration position simulation, γ B Represents the detail coefficient weight of the direction corresponding to matrix B, represents the square of the Euclidean norm, u B represents the mean of each element of matrix B, which belongs to the set of horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image at the mth calibration position, cA m , cH m ,cV m and cD m They respectively represent the approximate coefficient matrix, horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image at the mth calibration position, m=1,2,…,M, where M is the number of calibration positions.

[0024] Preferably, the calculation formula for weighted updating of the current axial distance is as follows:

[0025]

[0026]

[0027] Among them, z represents the updated axial distance, z 0 represents the axial distance to be calibrated, z′ represents the axial distance before update, δz represents the current axial distance update step, δz′ represents the axial distance update step obtained in the previous iteration, η represents the axial distance update step damping coefficient, dz m represents the mth propagation distance, It represents the clarity of the image of the sample to be tested simulated at the mth calibration position, m=1, 2, ..., M, where M is the number of calibration positions.

[0028] Preferably, the updating of the reconstructed sample to be tested at the current axial distance is specifically as follows:

[0029] (1) Using the current axial distance, calculate the new sample single pixel size:

[0030]

[0031] Among them, x d represents the single pixel size of the sample, λ represents the wavelength of the light source used, z represents the updated axial distance, and D represents the size of the detector target surface;

[0032] (2) using a single pixel size as a ratio and according to the characteristics of the coherent diffraction imaging system used, converting the actual scanning position coordinates into discrete scanning position coordinates;

[0033] (3) Substituting the discrete scanning position coordinates in the new sample matrix into the coherent diffraction imaging algorithm and iterating several times to obtain the reconstructed sample and illumination probe at the current axial distance.

[0034] Preferably, the propagation model in the coherent diffraction imaging algorithm is the same as or different from the near-field propagation model.

[0035] Preferably, the coherent diffraction imaging system is a transmission coherent diffraction imaging system or a reflection coherent diffraction imaging system.

[0036] It should be noted that the present invention is applicable to different types of coherent diffraction imaging systems and has strong applicability. The present invention aims to solve the problem of axial distance error between the sample to be measured and the detector in the coherent diffraction imaging reconstruction process. The traditional coherent diffraction imaging system, stacked diffraction imaging system, coherent modulation imaging system, multi-distance coherent diffraction imaging system and other systems developed based on coherent diffraction imaging have the problem of difficulty in accurately measuring the axial distance between the sample and the detector, so they are all applicable.

[0037] To achieve the above-mentioned object, in a second aspect, the present application provides a coherent diffraction imaging automatic focusing system, comprising at least one processor and at least one memory;

[0038] The at least one memory is used to store computer instructions;

[0039] The at least one processor is used to execute at least part of the computer instructions to implement the automatic focusing method described in the first aspect.

[0040] It can be understood that the beneficial effects of the second aspect mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here.

[0041] In general, the above technical solutions conceived by this application have the following beneficial effects compared with the prior art:

[0042] This application proposes a coherent diffraction imaging automatic focusing method, which simulates multiple groups of samples to be tested at different distances through a near-field propagation model, performs wavelet transformation on the simulated samples to be tested to obtain wavelet coefficients, calculates clarity based on the wavelet coefficients, and uses the clarity of all simulated samples to be tested at the current axial distance to weight and update the axial distance, so as to gradually approach the actual value of the axial distance between the sample and the detector, realize automatic focusing of stacked diffraction imaging, and then realize high-quality reconstruction of the sample to be tested and the illumination probe. This application proposes for the first time to use operators in the frequency domain to calculate the clarity of the reconstructed image. This clarity calculation method has higher axial contrast and noise robustness, and greatly improves the convergence speed, convergence accuracy and initial error robustness of the focusing algorithm; this application uses wavelet transform to extract detail components of the image in the horizontal, vertical and diagonal directions in the frequency domain, so as to construct an evaluation method for the clarity of the reconstructed image, and the weights of different directions can be flexibly adjusted according to needs, which has better generalization for samples with different patterns. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a flow chart of a coherent diffraction imaging automatic focusing method provided by this application.

[0044] Figure 2 1 is a schematic diagram of the optical path of a stacked diffraction computational imaging system provided in an embodiment of the present application, wherein (a) is a transmission-type stacked diffraction imaging system and (b) is a reflection-type stacked diffraction imaging system.

[0045] Figure 3 It is a flow chart of the stacked diffraction imaging automatic focusing method provided in an embodiment of the present application.

[0046] Figure 4 It is a pattern of the sample to be tested and the illumination probe used in the simulation process provided in the embodiment of the present application, wherein (a) is the amplitude information of the sample to be tested, (b) is the phase information of the sample to be tested, (c) is the amplitude information of the illumination probe, and (d) is the phase information of the illumination probe.

[0047] Figure 5 It is the reconstruction result without automatic focusing provided by the embodiment of the present application, wherein (a) is the reconstruction result of the amplitude of the sample to be tested without automatic focusing, (b) is the local enlarged image of the 1-3 groups of line pairs in (a), (c) is the reconstruction result of the amplitude of the illumination probe without automatic focusing, and (d) is the reconstruction result of the phase of the illumination probe without automatic focusing.

[0048] Figure 6It is the reconstruction result after automatic focusing provided in the embodiment of the present application, wherein (a) is the amplitude reconstruction result of the sample to be tested after automatic focusing, (b) is the local enlarged image of the line pair area of ​​groups 1-3 in (a), (c) is the amplitude reconstruction result of the illumination probe after automatic focusing, and (d) is the phase reconstruction result of the illumination probe after automatic focusing.

[0049] Figure 7 It is the axial distance convergence curve of the sample under different axial distance initial values ​​provided in the embodiment of the present application.

[0050] Throughout the drawings, the same reference numerals are used to denote the same elements or structures, wherein:

[0051] 1- He-Ne laser, 2- beam expander, 3- adjustable aperture, 4- focusing lens / focusing mirror, 5- sample to be measured, 6- detector. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0053] The term "and / or" in this article is a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The symbol " / " in this article indicates that the associated objects are in an or relationship, for example, A / B means A or B.

[0054] The terms "first" and "second" in the specification and claims herein are used to distinguish different objects rather than to describe a specific order of the objects. For example, a first response message and a second response message are used to distinguish different response messages rather than to describe a specific order of the response messages.

[0055] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0056] In the description of the embodiments of the present application, unless otherwise specified, "multiple" means two or more than two. For example, multiple processing units refer to two or more processing units, etc.; multiple elements refer to two or more elements, etc.

[0057] Next, the technical solutions provided in the embodiments of the present application are introduced.

[0058] like Figure 1 As shown, the present application provides a coherent diffraction imaging automatic focusing method, comprising:

[0059] S1. Obtain the axial distance to be calibrated and the reconstructed sample to be tested and the illumination probe at this distance;

[0060] S2. Using the near-field propagation model, the reconstructed sample to be tested at the current axial distance is propagated for several distances to obtain simulated samples to be tested at different calibration positions;

[0061] S3. Perform wavelet transform on each simulated sample to be tested to obtain corresponding wavelet coefficients, wherein the wavelet coefficients include approximation coefficients, horizontal detail coefficients, vertical detail coefficients and diagonal detail coefficients;

[0062] S4. Calculate the variance of the wavelet coefficients of each simulated sample to be tested and sum them to obtain the corresponding clarity;

[0063] S5. Using the clarity of all simulated samples to be tested as weights, weighted update of the current axial distance, and update of the reconstructed sample to be tested and the illumination probe at the current axial distance;

[0064] S6. Repeat S2-S5 until the current axial distance approaches the actual value of the axial distance, and use the current axial distance as the calibrated axial distance to achieve automatic focusing of coherent diffraction imaging, while obtaining a clear reconstruction of the sample to be tested and the illumination probe after calibration.

[0065] Preferably, step S2 is specifically as follows:

[0066] S21. Obtain the current axial distance and calibration step, set multiple calibration positions along the axial direction with the current axial distance as the center according to the calibration step, and obtain the propagation distance of each calibration position relative to the center point;

[0067] S22. Use the near-field propagation model to propagate the sample to be tested reconstructed at the current axial distance to different distances to simulate the sample to be tested reconstructed by the coherent diffraction imaging algorithm at different calibration positions.

[0068] Preferably, the calibration step is an integer multiple of the depth of field of the coherent diffraction system.

[0069] Preferably, step S3 is specifically as follows:

[0070] [cA mm , cH mm , cV mm , cD m ]=dwt2(O m )

[0071] Among them, cA mm , cH mm ,cV mm and cD mm denote the approximate coefficient matrix, horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image at the mth calibration position, dwt2 denotes a two-dimensional discrete wavelet transform function, O mm It indicates that the mth calibration position simulates the sample image to be tested, m=1, 2, ..., M, where M is the number of calibration positions.

[0072] Preferably, the calculation formula of the clarity is as follows:

[0073]

[0074] in, represents the clarity of the image of the sample to be tested at the mth calibration position simulation, γ B Represents the detail coefficient weight of the direction corresponding to matrix B, represents the square of the Euclidean norm, u B represents the mean of each element of matrix B, which belongs to the set of horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image at the mth calibration position, cA m , cH m ,cV m and cD m They respectively represent the approximate coefficient matrix, horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image at the mth calibration position, where m=1, 2, ..., M, where M is the number of calibration positions.

[0075] Preferably, the calculation formula for weighted updating of the current axial distance is as follows:

[0076]

[0077]

[0078] Among them, z represents the axial distance before updating, z 0 represents the axial distance to be calibrated, z′ represents the updated axial distance, δz represents the current axial distance update step, δz′ represents the axial distance update step obtained in the previous iteration, η represents the axial distance update step damping coefficient, dz mm represents the mth propagation distance, It represents the clarity of the image of the sample to be tested simulated at the mth calibration position, m=1, 2, ..., M, where M is the number of calibration positions.

[0079] Preferably, the updating of the reconstructed sample to be tested at the current axial distance is specifically as follows:

[0080] (1) Using the current axial distance, calculate the new sample single pixel size:

[0081]

[0082] Among them, x d represents the single pixel size of the sample, λ represents the wavelength of the light source used, z represents the current axial distance, and D represents the size of the detector target surface;

[0083] (2) using a single pixel size as a ratio and according to the characteristics of the coherent diffraction imaging system used, converting the actual scanning position coordinates into discrete scanning position coordinates;

[0084] (3) Substituting the discrete scanning position coordinates in the new sample matrix into the coherent diffraction imaging algorithm and iterating several times to obtain the reconstructed sample and illumination probe at the current axial distance.

[0085] Preferably, the propagation model in the coherent diffraction imaging algorithm is the same as or different from the near-field propagation model.

[0086] Preferably, the conventional coherent diffraction imaging system, the coherent modulation imaging system, the holographic imaging system, the transmission-type stacked diffraction imaging system and the reflection-type stacked diffraction imaging system.

[0087] Example

[0088] The present application is applicable to different types of coherent diffraction imaging systems, including but not limited to: traditional coherent diffraction imaging systems, coherent modulation imaging systems, multi-distance coherent diffraction imaging systems, transmission stacked diffraction imaging systems and reflection stacked diffraction imaging systems.

[0089] In this embodiment, the coherent diffraction imaging system adopts a stacked diffraction imaging system, and the near-field propagation model adopts an angular spectrum propagation model.

[0090] This embodiment provides a stacked diffraction imaging method, which is applied to a stacked diffraction computational imaging system, including:

[0091] Step S1: Building a stacked diffraction imaging system.

[0092] Figure 2 1 is a schematic diagram of the optical path of the stacked diffraction computational imaging system provided in the embodiment of the present application, wherein (a) is a transmission-type stacked diffraction imaging system, and (b) is a reflection-type stacked diffraction imaging system. The stacked diffraction system comprises a He-Ne laser 1, a beam expander 2, an adjustable aperture 3, a focusing lens / focusing reflector 4, a sample to be measured 5, and a detector 6, which are sequentially arranged along the optical path.

[0093] Step S2: Adjust the optical path system, the HeNe laser 1 emits a working wavelength of 632.8nm and a beam diameter (1 / e 2 ) is a 0.54mm light beam, which is expanded 10 times by the beam expander 2 and propagated to the adjustable aperture 3. The light beam size is adjusted to a suitable size by the adjustable aperture 3. The collimated parallel light beam passes through the focusing lens / focusing reflector 4 and irradiates the sample 5 to be tested. The sample 5 to be tested is placed near the rear focal plane of the focusing lens / focusing reflector 4. The position and angle of the detector 6 are adjusted so that the zero-order diffraction light is in the center area of ​​the detector target surface and the detector target surface is perpendicular to the optical axis. According to the actual experimental needs, the size of the adjustable aperture 3 is adjusted and the sample 5 to be tested is moved along the optical axis to adjust the specific light beam size. Generally, the diameter of the illumination spot irradiated on the sample 5 to be tested is 0.2mm~2mm.

[0094] Step S3: The precision motion stage drives the sample 5 to move in the plane of the sample according to the pre-generated scanning trajectory (including but not limited to grating curves, concentric circle curves, Fermat curves, etc.), while ensuring that there is an overlap area of ​​more than 60% between adjacent illumination spots, that is, the spatial overlap constraint is satisfied. Record the scanning position coordinates r j , where j (j=1,2,…,J) is the serial number of the corresponding scanning position.

[0095] Step S4: The detector 6 at a certain distance from the sample 5 records the intensity information I of a series of diffracted light fields generated after the illumination spot at each scanning position on the motion trajectory interacts with the sample 5. j (f x ,f y ), where (f x ,f y ) is the frequency domain coordinate. The approximate axial distance z between the sample 5 to be tested and the target surface of the detector 6 is measured. 0 .

[0096] Step S5: The intensity information I of the diffracted light field measured in step S4 is j (f x ,f y ) and the distance z between the sample and the detector target 0 Substituting into the stacked diffraction imaging algorithm, the uncalibrated reconstructed illumination probe complex amplitude function P is obtained. 0 (x, y) and the complex amplitude function O of the sample to be tested 0 (x,y), where (x,y) are spatial coordinates.

[0097] The detailed steps of using the stacked diffraction imaging algorithm to obtain the uncalibrated reconstructed illumination probe and the complex amplitude of the sample to be tested in step S5 are as follows:

[0098] S5.1: Use the axial distance z between the sample and the detector surface measured in S4 0 , calculated according to formula (1), the jth actual scanning position coordinate r is obtained j The corresponding discrete scanning position coordinates p in the sample matrix j :

[0099]

[0100] Among them, λ is the wavelength of the light source used, D is the size of the detector target surface, and round() is the rounding function.

[0101] S5.2: Make an initial guess for the complex amplitude function P(x,y) of the illumination probe and the complex amplitude function O(x,y) of the sample to be tested. Among them, Gaussian spots, circular spots with an intensity of 1, circular spots with random intensities, etc. can be used as the initial guess of the probe; all-1 matrices, random matrices, etc. can be used as the initial guess of the sample to be tested.

[0102] S5.3: For the jth scanning position, dot product the complex amplitude function P(x,y) of the illumination probe and the corresponding discrete scanning position coordinate p of the sample to be tested j The complex amplitude function O j (x,y), get the exit wave

[0103]

[0104] S5.4: Using the diffraction propagation model, the exit wave Propagate to the detector plane, and get the estimated diffraction light field distribution Φ of the detector plane j (f x ,f y ):

[0105]

[0106] Wherein, F[] is the propagation mode of the light field in free space, including but not limited to Fresnel propagation, Fraunhofer propagation, angular spectrum propagation, etc.

[0107] S5.5: retain the guessed diffraction light field distribution phase of the detector plane, and use the diffraction field intensity information I corresponding to the j-th scanning position obtained in step S4 j (f x ,f y ) replaces the guessed diffraction light field distribution amplitude of the detector plane, and obtains the new diffraction light field distribution of the detector plane Φ j '(f x ,f y ):

[0108]

[0109] S5.6: The new detector plane diffracted light field distribution Φ j '(f x ,f y ) is propagated back to the sample plane to obtain the updated exit wave

[0110]

[0111] Among them, F[] is the reverse propagation mode of the light field in free space.

[0112] S5.7: Based on the exit waves before and after the update obtained in S5.3 and S5.6, the complex amplitude functions P'(x, y) and O of the illumination probe and the sample to be measured are updated using the stacked diffraction imaging algorithm. j '(x,y):

[0113]

[0114] Among them, α and β are the update steps of the illumination probe and the sample to be tested, respectively, and generally take values ​​between [0,1].

[0115] S5.8: Randomly select the next scanning position and repeat S5.3-S5.7 until all J positions are updated.

[0116] S5.9: Calculate the root mean square error between the simulated diffraction light field distribution and the measured diffraction light field distribution at each scanning position. When the root mean square error is less than the set threshold or reaches the preset upper limit of the number of iterations, output the complex amplitude function P'(x, y) and O'(x, y) of the illumination probe and the sample to be tested at this time as the uncalibrated illumination probe P 0 (x, y) and the sample to be tested O 0 (x, y), otherwise, go to step S5.3.

[0117] Step S6: The intensity information I of the light field measured in step S4 is j (f x ,f y ) and the distance z between the sample and the detector target 0 , the preliminary reconstructed illumination probe complex amplitude function P obtained in step S5 0 (x, y) and the complex amplitude function O of the sample to be tested 0 (x, y) is substituted into the following stacked diffraction imaging autofocus method, and the complex amplitude function P(x, y) of the illumination probe and the complex amplitude function O(x, y) of the sample to be measured are reconstructed at the same time, and the axial distance z is calibrated.

[0118] Figure 3: is a flow chart of the stacked diffraction imaging automatic focusing method provided in the embodiment of the present application. Figure 3 As shown, the specific process of step S6 is as follows:

[0119] S6.1: With the axial distance z' to be calibrated as the center, set a total of M = 2n + 1 calibration positions with a sampling interval Δz, and calculate the axial distance dz of the mth (m = 1, 2, ..., M) calibration position relative to the center m , preferably, set Δz = i × Dof, i is generally an integer between [1,5], and n is generally an integer between [20,60].

[0120] Wherein, Dof is the depth of field of the stacked diffraction imaging system, which is calculated according to formula (8):

[0121]

[0122] Among them, NA is the numerical aperture of the system, which is determined by the size of the detector target surface and the distance between the detector target surface and the sample.

[0123] S6.2: For the mth propagation distance dz m , using the angular spectrum propagation model to propagate the sample O(x,y) to be tested dz m , the simulated axial distance is z'+dz m The sample to be tested at position O m (x,y):

[0124]

[0125] in, represents the Fourier transform, Represents the inverse Fourier transform.

[0126] S6.3: The simulated sample O obtained in step S6.2 m (x, y) is transformed into a two-dimensional discrete wavelet transform to obtain the wavelet coefficient matrix of the simulated sample to be tested:

[0127] [cA m ,cH m ,cV m ,cD m ]=dwt2(O m )(10)

[0128] Among them, cA m , cH m ,cV m and cD mThe approximate coefficient matrix, horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the sample image to be tested are simulated for the mth propagation position respectively, and dwt2 is a two-dimensional discrete wavelet transform function. Preferably, the haar wavelet is selected as the wavelet basis.

[0129] S6.4: Apply the approximate coefficient matrix, the horizontal detail coefficient matrix, the vertical detail coefficient matrix and the diagonal detail coefficient matrix obtained in step S6.3 to calculate the sharpness of the simulated sample to be tested at the mth axial position

[0130]

[0131] Among them, γ B Represents the detail coefficient weight of the direction corresponding to matrix B, is the square of the Euclidean norm, u B is the mean of the elements of matrix B, and matrix B belongs to the set consisting of the horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image to be tested at the mth propagation position.

[0132] S6.5: Select the next axial position and repeat steps S6.2-S6.4 until the clarity evaluation values ​​of all M axial positions are obtained.

[0133] S6.6: Based on the axial propagation distance dz set in S6.1 m and the clarity obtained in S6.5 Calculate the axial distance update step δz:

[0134]

[0135] Among them, δz′ is the axial distance update step obtained in the previous iteration, η is the axial distance update step damping coefficient, which is generally taken between [0,1], and dz m represents the mth propagation distance.

[0136] S6.7: Update the axial distance according to the axial distance update step δz obtained in S6.6:

[0137] z=z′-δz (13)

[0138] Wherein, z is the updated axial distance, and z' is the axial distance before the update.

[0139] S6.8: Update the single pixel size x of the sample matrix according to the new axial distance z obtained in S6.7 d :

[0140] x d =λz / D (14)

[0141] S6.9: Calculate the new actual scanning coordinate r according to the single pixel size of the sample matrix obtained in S6.8 j The corresponding discrete scan coordinate position p in the sample matrix j :

[0142] p j =round(r j / x d ) (15)

[0143] S6.10: Based on the discrete scan coordinate position p in the sample matrix updated in S6.9 j , using the stacked diffraction iterations of steps S5.3-S5.8 for k times, the complex amplitude function P(x, y) of the illumination probe and the complex amplitude function O(x, y) of the sample to be measured at the updated axial distance are obtained. Preferably, k is set to 5.

[0144] S6.11: Calculate the variation interval ΔZ of the axial distance z in the past n iterations. When ΔZ is less than the set threshold or the number of iterations reaches the preset upper limit, output the last updated axial distance as the calibrated axial distance z, and output the complex amplitude function P(x, y) of the illumination probe and the complex amplitude function O(x, y) of the sample under this axial distance as the calibrated illumination probe and the sample under test, respectively. Otherwise, set the updated axial distance z as the axial distance z' to be calibrated, and enter step S6.1.

[0145] Figure 4 is a diagram of the sample to be tested and the illumination probe used in the simulation process provided in the embodiment of the present application. The sample to be tested used in the simulation experiment has amplitude information as follows Figure 4 The size shown in (a) is a 512×512 pixel resolution target, which is expanded to 768×768 pixels by padding the edges with 0 to ensure that the full resolution target can be reconstructed. The phase information is as follows: Figure 4 As shown in (b), a Gaussian beam with a wavelength of 632.8nm and a beam diameter of 128Pixel is constructed. After passing through a focusing lens with a focal length of 2.5mm, it propagates 5mm to the plane of the sample to be tested as an illumination probe. Its amplitude information is as follows Figure 4 As shown in (c), the phase information is Figure 4As shown in (d). The exit wave formed by the interaction between the illumination probe and the sample propagates 50 mm before reaching the detector target surface, and the detector target surface size is 256×256Pixel. The precision motion stage drives the sample to be tested to move along a preset trajectory in the plane perpendicular to the optical axis. The set trajectory is a 17×17 grid path, in which the interval between two adjacent points is about 24Piexl. Each point in the grid is set with a certain random offset to avoid periodic artifacts. The detector collects the diffraction field data generated by all 289 different positions.

[0146] Figure 5 This is the reconstruction result without autofocusing provided by the embodiment of the present application, wherein (a) is the reconstruction result of the amplitude of the sample to be tested without autofocusing, (b) is the local enlarged view of the line pair area of ​​1-3 groups in (a), (c) is the reconstruction result of the amplitude of the illumination probe without autofocusing, and (d) is the reconstruction result of the phase of the illumination probe without autofocusing. The axial distance between the sample to be tested and the detector target surface is set to 55mm, a random matrix is ​​used as the initial guess of the sample to be tested, a circular spot with an intensity of 1 is used as the initial guess of the illumination probe, and the stacked diffraction imaging algorithm (mPIE) is iterated 400 times. The simulation results are shown in FIG. Figure 5 From the simulation results, it can be seen that the reconstructed amplitude of the sample to be tested is relatively fuzzy, and there are ghost images in the 1-3 groups of line pairs.

[0147] Figure 6 The reconstruction result after autofocusing provided in the embodiment of the present application, wherein (a) is the amplitude reconstruction result of the sample to be tested after autofocusing, (b) is the local enlarged view of the line pair area of ​​1-3 groups in (a), (c) is the amplitude reconstruction result of the illumination probe after autofocusing, and (d) is the phase reconstruction result of the illumination probe after autofocusing. The initial axial distance between the sample to be tested and the detector target surface is set to 55mm, a random matrix is ​​used as the initial guess of the sample to be tested, a circular spot with an intensity of 1 is used as the initial guess of the illumination probe, and the above-mentioned stacked diffraction imaging autofocusing algorithm is iterated 400 times. The simulation results are as follows: Figure 6 From the simulation results, it can be seen that the reconstructed sample to be tested is very clear and there is no ghosting in the reconstruction result.

[0148] Figure 7The axial distance convergence curve of the sample under different initial axial distance values ​​provided in the embodiment of the present application. The initial axial distance between the sample to be tested and the detector target surface is set to 45, 47.5, 49, 51, 52.5, and 55 mm respectively, a random matrix is ​​used as the initial guess of the sample to be tested, a circular spot with an intensity of 1 is used as the initial guess of the illumination probe, and the above-mentioned stacked diffraction imaging automatic focusing algorithm is iterated 400 times to record the axial distance calibration curve under different initial axial distances. It can be seen from the curve that under different initial axial distances, after 190 iterations of the focusing algorithm, the axial distance converges to 50±0.005mm.

[0149] The simulation results show that without automatic focusing, the correspondence between the single pixel size on the detector target surface and the single pixel size in the sample matrix is ​​incorrect, causing the sample to be reconstructed by the stacked diffraction imaging algorithm to be scaled and blurred, and even ghosting, which reduces the system resolution. After focusing by the automatic focusing algorithm, the axial distance is calibrated to near the true value, eliminating the ghosting in the reconstruction result, making the reconstruction result clear. Figure 7 It is shown that for an initial axial distance error of up to 5 mm, the coherent diffraction imaging autofocusing method proposed in this application can quickly converge to the true value, demonstrating its strong robustness to the initial axial distance error.

[0150] It can be understood that the detailed functional implementation of each of the above-mentioned units / modules can be found in the introduction of the aforementioned method embodiment, and will not be repeated here.

[0151] It should be understood that the above-mentioned device is used to execute the method in the above-mentioned embodiment. The implementation principle and technical effect of the corresponding program module in the device are similar to those described in the above-mentioned method. The working process of the device can refer to the corresponding process in the above-mentioned method, which will not be repeated here.

[0152] Based on the method in the above embodiment, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method in the above embodiment.

[0153] Based on the method in the above embodiment, an embodiment of the present application provides a computer program product. When the computer program product runs on a processor, the processor executes the method in the above embodiment.

[0154] It is understandable that the processor in the embodiment of the present application can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. The general-purpose processor can be a microprocessor or any conventional processor.

[0155] The method steps in the embodiments of the present application can be implemented by hardware or by a processor executing software instructions. The software instructions can be composed of corresponding software modules, and the software modules can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, mobile hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an ASIC.

[0156] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions may be transmitted from a website site, a computer, a server or a data center to another website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state disk (SSD)), etc.

[0157] It should be understood that the various numerical numbers involved in the embodiments of the present application are only used for the convenience of description and are not used to limit the scope of the embodiments of the present application.

[0158] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A coherent diffraction imaging automatic focusing method, characterized in that: include: S1. Obtain the axial distance to be calibrated and the reconstructed sample to be tested and the illumination probe at this distance; S2. Using the near-field propagation model, the reconstructed sample to be tested at the current axial distance is propagated for several distances to obtain simulated samples to be tested at different calibration positions; S3. Perform wavelet transform on each simulated sample to be tested to obtain corresponding wavelet coefficients, wherein the wavelet coefficients include approximation coefficients, horizontal detail coefficients, vertical detail coefficients and diagonal detail coefficients; S4. Calculate the variance of the wavelet coefficients of each simulated sample to be tested and sum them to obtain the corresponding clarity; S5. Using the clarity of all simulated samples to be tested as weights, weighted update of the current axial distance, and update of the reconstructed sample to be tested and the illumination probe at the current axial distance; S6. Repeat S2-S5 until the current axial distance approaches the actual value of the axial distance, and use the current axial distance as the calibrated axial distance to achieve automatic focusing of coherent diffraction imaging, while obtaining a clear reconstruction of the sample to be tested and the illumination probe after calibration.

2. The automatic focusing method according to claim 1, wherein: Step S2 is specifically as follows: S21. Obtain the current axial distance and calibration step, set multiple calibration positions along the axial direction with the current axial distance as the center according to the calibration step, and obtain the propagation distance of each calibration position relative to the center point; S22. Use the near-field propagation model to propagate the sample to be tested reconstructed at the current axial distance to different distances to simulate the sample to be tested reconstructed by the coherent diffraction imaging algorithm at different calibration positions.

3. The automatic focusing method according to claim 2, wherein: The calibration step is an integer multiple of the depth of field of the coherent diffraction system.

4. The automatic focusing method according to claim 1, wherein: Step S3 is as follows: [cA m ,cH m ,cV m ,cont'd m ]=dwt2(O m ) Among them, cA m , cH m ,cV m and cD m denote the approximate coefficient matrix, horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image at the mth calibration position, dwt2 denotes a two-dimensional discrete wavelet transform function, O m It indicates that the mth calibration position simulates the sample image to be tested, m = 1, 2, ..., M, where M is the number of calibration positions.

5. The automatic focusing method according to claim 1, wherein: The calculation formula of the clarity is as follows: in, represents the clarity of the image of the sample to be tested at the mth calibration position simulation, γ B Represents the detail coefficient weight of the direction corresponding to matrix B, represents the square of the Euclidean norm, u B represents the mean of each element of matrix B, which belongs to the set of horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image at the mth calibration position, cA m , cH m ,cV m and cD m They respectively represent the approximate coefficient matrix, horizontal detail coefficient matrix, vertical detail coefficient matrix and diagonal detail coefficient matrix of the simulated sample image at the mth calibration position, m=1,2,…,M, where M is the number of calibration positions.

6. The automatic focusing method according to claim 1, wherein: The calculation formula for weighted updating of the current axial distance is as follows: Where z represents the updated axial distance, z0 represents the axial distance to be calibrated, z′ represents the axial distance before the update, δz represents the current axial distance update step, δz′ represents the axial distance update step obtained in the previous iteration, η represents the axial distance update step damping coefficient, dz m represents the mth propagation distance, It indicates the clarity of the image of the sample to be tested simulated at the mth calibration position, m = 1, 2, ..., M, where M is the number of calibration positions.

7. The automatic focusing method according to claim 1, wherein: The updating of the reconstructed sample to be tested at the current axial distance is specifically as follows: (1) Using the current axial distance, calculate the new sample single pixel size: Among them, x d represents the single pixel size of the sample, λ represents the wavelength of the light source used, z represents the updated axial distance, and D represents the size of the detector target surface; (2) using a single pixel size as a ratio and according to the characteristics of the coherent diffraction imaging system used, converting the actual scanning position coordinates into discrete scanning position coordinates; (3) Substituting the discrete scanning position coordinates in the new sample matrix into the coherent diffraction imaging algorithm and iterating several times to obtain the reconstructed sample and illumination probe at the current axial distance.

8. The automatic focusing method according to claim 1, wherein: The propagation model in the coherent diffraction imaging algorithm is the same as or different from the near-field propagation model.

9. The automatic focusing method according to any one of claims 1 to 8, characterized in that: The coherent diffraction imaging system is a transmission-type coherent diffraction imaging system or a reflection-type coherent diffraction imaging system.

10. A coherent diffraction imaging automatic focusing system, characterized in that: comprising at least one processor and at least one memory; The at least one memory is used to store computer instructions; The at least one processor is configured to execute at least part of the computer instructions to implement the automatic focusing method according to any one of claims 1 to 9.

Images