A high dynamic range image fusion lamination diffraction imaging method and system

Abstract

The application belongs to the field of laminated diffraction imaging, and specifically discloses a high dynamic range image fusion laminated diffraction imaging method and system, the imaging method being: initializing an illumination probe and a sample to be measured, constructing a joint distribution containing a diffraction light field, a dark field and an exposure time sequence, solving a maximum likelihood estimation of absolute irradiance based on a joint probability density function, fusing a diffraction field high dynamic range absolute irradiance from directly measured diffraction light field, dark field and exposure time parameters, replacing an analog diffraction light field amplitude and keeping its phase unchanged, and simultaneously reconstructing the illumination probe and the sample to be measured according to the analog diffraction light field before and after updating. The application considers the influence of mixed noise, effectively extracts high-frequency overlapping correlation signals without losing or reducing the signal-to-noise ratio of low-frequency region diffraction signals, significantly reduces the requirement for the dynamic range of a detector, and avoids the increase of imaging system uncertainty without additional complex parameter calibration and modulation equipment in the imaging process.

Description

Technical Field

[0001] This invention belongs to the field of stacked diffraction imaging technology, and more specifically, relates to a high dynamic range image fusion stacked diffraction imaging method and system. Background Technology

[0002] Stacked diffraction imaging is a lens-free and interference-free computational imaging technique. By acquiring the diffraction light field information of the interaction between the illumination probe and the sample, and utilizing the real-domain spatial overlap constraint and the frequency-domain spatial light intensity constraint, the complex amplitude information of the illumination probe and the sample is iteratively reconstructed using the stacked diffraction algorithm. It has been widely used in emerging nanomaterials, semiconductor physics and cell biology.

[0003] The Abbe diffraction limit indicates that the resolution of an imaging system depends only on the numerical aperture (NA) and the illumination wavelength. In wavelength-invariant stacked diffraction imaging, the resolution depends entirely on the effective NA, which is the angular range from weak high-frequency scattering unaffected by noise to strong low-frequency scattering in Fourier space. In reciprocal space, the scattering intensity in the near-zero angle low-scattering vector region is several orders of magnitude greater than that in the high-scattering vector region. However, for pixel charges with full-well effects, the detector only records the coherent diffraction field within a certain dynamic range, severely affecting the signal-to-noise ratio of diffraction fields with different photon fluxes, and thus impacting the effective numerical aperture and resolution of diffraction imaging.

[0004] Currently, techniques to enhance the resolution of an effective photodetector (NA) by widening its dynamic range have been extensively studied. In the field of lensless diffraction imaging, there are two main research routes. The first approach involves optimizing the experimental structure and improving the experimental setup to reduce the ratio of diffraction field intensity between low-scattering and high-scattering vector regions within a limited dynamic range. This includes beam blocking and optical path modulation. Methods such as placing beam blocks in the detector's central plane to reduce non-ideal background scattering, designing special masks to modulate the incident wave, or introducing reference objects and apertures to modulate the outgoing wave enhance imaging resolution. While this method introduces modulators to reduce the dynamic range requirement of the detector, it increases the complexity of the imaging system and amplifies system uncertainty and instability. The second approach emphasizes acquiring a high dynamic range diffraction field that exceeds the detector's original limited dynamic range, including methods such as multiple exposure fusion. In the field of stacked diffraction imaging, this is mainly achieved by calibrating the linear response function of the photoelectric sensor to obtain a high dynamic range image through multiple exposure fusion. However, in this method, the actual diffracted light field inevitably couples with detector noise, which will affect the accuracy of the final fused and reconstructed image. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the present invention aims to provide a high dynamic range image fusion stacked diffraction imaging method and system, which solves the problems of existing methods requiring additional modulation devices or masks, thereby increasing system uncertainty and instability, as well as the problems of mixed noise affecting the process of existing image fusion methods and pixel culling operations causing the loss of diffraction information.

[0006] To achieve the above objectives, according to one aspect of the present invention, a high dynamic range image fusion stacked diffraction imaging method is provided, comprising the following steps:

[0007] S1 initializes the illumination probe and the sample to be tested, acquires the scanning position information, diffraction light field intensity information and dark field information of the sample to be tested at different exposure times, and synthesizes high dynamic range irradiance based on the exposure time, diffraction light field intensity information and dark field information.

[0008] The S2 simulates the first emitted light generated by the interaction between the illumination probe and the sample at different positions along the scanning path. The first emitted light is then propagated to the detector plane through a two-dimensional scalar diffraction propagation model to obtain a simulated diffraction field. This simulated diffraction field is then used to simulate the detector to measure the first emitted light.

[0009] S3 replaces the intensity of the simulated diffraction field with high dynamic range irradiance while keeping its phase unchanged to obtain an updated simulated diffraction field; the updated simulated diffraction field is then propagated in reverse using a scalar diffraction inverse propagation model to obtain an updated second outgoing light.

[0010] S4 Based on the first and second emitted light, the illumination probe and the sample to be tested are updated simultaneously, and the updated illumination probe and the sample to be tested are used as the initial guess for the next position. Steps S2-S3 are repeated until all scanning positions are updated.

[0011] S5 calculates the root mean square error of the intensity of the simulated diffraction field and the high dynamic range irradiance at all scanning positions. When the root mean square error is less than a preset threshold, the iterated illumination probe and the sample to be tested are output; otherwise, proceed to step S2.

[0012] Furthermore, in step S1, the method for synthesizing high dynamic range irradiance includes the following steps:

[0013] S101 constructs a joint distribution containing diffraction light field information, dark field information, and exposure time series at each scanning position, and obtains the joint probability density function corresponding to the joint distribution;

[0014] S102 At each scanning position, a high dynamic range absolute irradiance is synthesized based on the joint probability density function;

[0015] Preferably, at any scan position, the constructed joint distribution is represented by the first formula:

[0016]

[0017] Among them, Q i For t i Joint distribution constructed by exposure time, I i B For after t i Exposure time diffraction field readout; I i D For after t i Exposure time dark field readout; k is the total gain coefficient during the conversion process; F is the absolute irradiance received by the detector; t i For the i-th exposure time; and The representatives are from I i B I i D Different independent distributions; U is the dark current shot noise from photon to charge, and A is the readout noise from charge to digital readout; and It follows a Poisson difference distribution, and approaches a Gaussian distribution when the absolute irradiance is large. They also follow a Gaussian distribution;

[0018] The expected value and variance of the joint distribution are calculated using the following formulas:

[0019]

[0020]

[0021] in and For t i The expected value and variance of the joint distribution of exposure times, i = 1, 2, ..., N, k is the total gain coefficient in the conversion process, ε F ε is the mathematical expectation of absolute irradiance. U For the mathematical expectation of shot noise, To read the variance of the noise.

[0022] Furthermore, in the first formula, the conditional probability equals the joint probability of the joint distribution; therefore, based on {Q} for each exposure time... i} i＝1,2,…,N The joint probability density function obtained from conditional probability is expressed as:

[0023]

[0024] Where, ε Q For t i Mathematical expectation of the joint distribution of exposure times; For t i The variance of the joint distribution of exposure times, i = 1, 2, ..., N; q i That is Q i , for t i Joint distribution under exposure time.

[0025] Furthermore, in step S102, the maximum likelihood estimate of the absolute irradiance is solved according to the joint probability density function, and then the high dynamic range absolute irradiance is synthesized based on the maximum likelihood estimate.

[0026] Preferably, the maximum likelihood estimate is obtained through a second formula:

[0027]

[0028] Among them, I i B For after t i Exposure time diffraction field readout; I i D For after t i Exposure time dark field readout; k is the total gain coefficient during the conversion process; F represents the absolute irradiance received by the detector; ε Q For ε Q q is the mathematical expectation of the joint distribution; i That is Q i , for t i Joint distribution under exposure time; For t i The variance of the joint distribution of exposure time; ε F This is a mathematical expectation for approximating absolute irradiance.

[0029] Furthermore, the synthesized high dynamic range absolute irradiance is expressed by the third formula:

[0030]

[0031] Wherein, the weight function w i Represented as: F HDR The mathematical expectation of absolute irradiance, which is an approximate estimate of absolute irradiance from diffraction field information and dark field information, is expressed as:

[0032]

[0033] Furthermore, the updated simulated diffraction field is represented as follows:

[0034]

[0035] Where, Φ j (q) represents the simulated diffraction field before the update in step S5, || denotes taking the elemental amplitude, F HDR,j This represents the high dynamic range irradiance at the j-th scan position.

[0036] Furthermore, in step S4, the fourth formula is used to simultaneously update the illumination probe and the sample to be tested:

[0037]

[0038]

[0039] Where α and β represent different iterative search step size factors, with values ​​between [0,1]; O j (r) represents the sample to be tested in the probe before the update at the j-th scan position, O' j P(r) represents the sample to be tested in the probe after the update at the j-th scan position; P(r) represents the illumination probe before the update, and P'(r) represents the illumination probe after the update. This represents the emitted light at the j-th scan position before the update. Represents the updated outgoing light at the j-th scan position; (r) represents the real space coordinates; * represents the conjugate operation of a complex matrix; max This represents the maximum value of each element in the matrix.

[0040] Furthermore, in step S6, the formula for calculating the root mean square error of the intensity of the simulated diffraction field and the high dynamic range irradiance at all scanning positions is expressed as follows:

[0041]

[0042] Among them, F HDR,j Φ represents the high dynamic range irradiance at the j-th scan position. j (q) represents the simulated diffraction field before the update in step S5, and MSE represents the mean square error.

[0043] Furthermore, the exposure time series is [2] 0 ,2 0.5 ,2 1 ,…,2 8 ]ms.

[0044] According to another aspect of the present invention, a stacked diffraction imaging system for implementing the high dynamic range image fusion stacked diffraction imaging method as described in any of the preceding claims is also disclosed, wherein the stacked diffraction imaging system is a transmission imaging system or a reflection imaging system.

[0045] Compared with the prior art, the above technical solutions conceived by this invention have the following main advantages:

[0046] (1) The high dynamic range image fusion stacked diffraction imaging method proposed in this invention has the advantages of being simple and easy to implement and having noise resistance and robustness. By constructing a joint distribution including the diffraction light field, dark field and exposure time series, the maximum likelihood estimate of the absolute irradiance is obtained. The high dynamic range absolute irradiance is obtained from a series of directly measured parameters (including diffraction light field, dark field and exposure time), which relaxes the requirements of the stacked diffraction imaging system on the high dynamic range of the detector and greatly improves the signal-to-noise ratio of the diffraction signal, the resolution and contrast of the reconstructed image.

[0047] (2) This invention takes into account the influence of mixed noise for the first time, and effectively extracts high-frequency overlapping correlation signals without losing or reducing the diffraction signal-to-noise ratio in the low-frequency region.

[0048] (3) Compared with the high dynamic range image fusion method based on linear response function correction, the present invention reduces the influence of noise on the diffraction signal before the phase recovery of stacked diffraction, and does not have any pixel removal operation, thus completely preserving the diffraction signal.

[0049] (4) The imaging method of the present invention is easy to implement and has better noise resistance. It can be widely used in fields such as digital holography, tomography, infrared imaging, fluorescence imaging and photoacoustic imaging, and has a wider range of applications.

[0050] (5) Compared with conventional beam blocking and modulation methods, the imaging system and imaging method of the present invention do not require the introduction of additional modulation equipment and complex parameter calibration process, thus avoiding the increase in the complexity and uncertainty of the imaging system. Attached Figure Description

[0051] Figure 1 This is a flowchart of the high dynamic range image fusion stacked diffraction imaging method provided in Embodiment 1 of the present invention;

[0052] Figure 2 This is a schematic diagram of the optical path of the stacked diffraction computational imaging system provided in Embodiment 1 of the present invention;

[0053] Figure 3 These are the amplitude and phase images of the sample under test and the illumination probe used in the simulation process provided in Embodiment 1 of the present invention;

[0054] Figure 4 It is the diffraction field fused from the single-exposure diffraction field and the multi-exposure high dynamic range image during the simulation process provided in Embodiment 1 of the present invention;

[0055] Figure 5These are the single-exposure reconstruction results and the multi-exposure high dynamic range image fusion reconstruction results provided in Embodiment 1 of the present invention;

[0056] Figure 6 This is the root mean square error curve of single-exposure reconstruction and multi-exposure high dynamic range image fusion reconstruction under different detector dynamic ranges provided in Embodiment 1 of the present invention.

[0057] In the figure: 1-Helium-neon laser, 2-Beam expander, 3-Adjustable aperture, 4-Focusing lens / focusing mirror, 5-Sample to be tested, 6-Sample to be tested. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0059] This invention provides a high dynamic range image fusion stacked diffraction imaging method, which can be applied to stacked diffraction computational imaging systems, combined with... Figure 1 As shown, the specific steps include:

[0060] Step S1: Construct a stacked diffraction imaging system, including but not limited to transmission imaging systems and reflection imaging systems.

[0061] like Figure 2 The diagram shows the optical path of the stacked diffraction computational imaging system provided in this embodiment. (a) is a transmission-type stacked diffraction computational imaging system, and (b) is a reflection-type stacked diffraction computational imaging system. The stacked diffraction system is configured as follows: helium-neon laser 1, beam expander 2, adjustable aperture 3, focusing lens / focusing mirror 4, sample to be tested 5, and detector 6.

[0062] Step S2: Adjust the optical path system to a working wavelength of 632.8 nm and a beam diameter of (1 / e) 2 The output beam of a 0.54mm helium-neon laser is expanded 5 times by a beam expander, and then the output beam diameter is adjusted to 1mm by an adjustable aperture. The collimated parallel beam is then focused onto the sample under test after passing through a focusing lens / focusing mirror with an effective focal length of 20cm. The sample under test is positioned approximately 1.5mm away from the back focal plane of the focusing mirror.

[0063] Step S3: The precision motion platform drives the sample to be tested to perform two-dimensional planar motion along a specific trajectory, while ensuring that adjacent illumination spots have sufficient overlapping area, i.e., spatial overlap constraint.

[0064] Step S4: The detector at a distance of 50 mm from the sample records a series of diffraction field readouts Z for different exposure times. i,j The subscript i indicates the exposure time t. i The subscript j indicates the j-th scan position.

[0065] Step S5: Read out the diffraction field values ​​Z for a series of different exposure times recorded in step S4. i,j The complex amplitude function O(r) of the sample under test and the complex amplitude function P(r) of the illumination probe are reconstructed by 250 iterations in the high dynamic range image fusion stacked diffraction imaging algorithm, where r is the spatial coordinate and the exposure time series is [2]. 0 ,2 0.5 ,2,...,2 8 ]ms.

[0066] The detailed steps of step S5 are as follows:

[0067] S5.1: Make initial guesses about the complex amplitude functions P(r) and O(r) of the illumination probe and the sample to be tested;

[0068] S5.2: At all scanning positions, construct a joint distribution containing the diffraction field, dark field, and exposure time series, calculate its expected value and variance, and obtain its joint probability density function; at one scanning position, its joint distribution is expressed as follows:

[0069]

[0070] Among them, Q i For t i Joint distribution constructed by exposure time, I i B For after t i Exposure time diffraction field readout; I i D For after t i Exposure time dark field readout; k is the total gain coefficient during the conversion process; F is the absolute irradiance received by the detector; t i For the i-th exposure time; and The representatives are from I i B I i D Different independent distributions; U is the dark current shot noise from photon to charge, and A is the readout noise from charge to digital readout; and It follows a Poisson difference distribution (Skellam distribution), and approaches a Gaussian distribution when the absolute irradiance is large. It also follows a Gaussian distribution.

[0071] The expected value and variance of the above joint distribution can be expressed as:

[0072]

[0073]

[0074] in and For t i Mathematical expectation and variance of the joint distribution of exposure times.

[0075] In the mathematical derivation above, the conditional probability equals the joint probability of the joint distribution, which is given by {Q} for each exposure time. i} i＝1,2,…,N The joint probability density function derived from conditional probability is:

[0076]

[0077] Repeat the same derivation process at all scan locations to calculate the corresponding joint probability density function.

[0078] S5.3: Solve for the maximum likelihood estimate of absolute irradiance based on the joint probability density function, and synthesize the high dynamic range absolute irradiance. The maximum likelihood estimate function of absolute irradiance is expressed as follows:

[0079]

[0080] Based on its first-order partial derivative, the maximum likelihood estimate of the absolute irradiance can be obtained as follows:

[0081]

[0082] With different gain coefficients k, the high dynamic range irradiance of signals with different coherence can be retrieved. The synthesized high dynamic range irradiance is given by the following formula:

[0083]

[0084] Wherein, the weight function w i as follows:

[0085]

[0086] The mathematical expectation of the absolute irradiance, approximated from the diffraction field information and the dark field information, is as follows:

[0087]

[0088] The same derivation process was repeated at all scanning locations to synthesize the corresponding high dynamic range irradiance.

[0089] S5.4: Following the scanning path preset in step S3, the initial guesses P(r) and O(r) of the illumination probe and the sample to be tested interact to form the emitted light function, which is expressed as the following formula:

[0090]

[0091] Among them, O j (r) represents the complex amplitude function of the sample under test within the probe at the j-th scanning position.

[0092] S5.5: The outgoing light function guessed in step S5.4 above propagates to the reciprocal space through the two-dimensional scalar diffraction propagation model, and the simulated diffraction light field of the detector plane is obtained as follows:

[0093]

[0094] Where, Φ j (q) represents the diffraction field of the emitted light at the j-th scanning position in reciprocal space; prop represents a two-dimensional scalar diffraction propagation model (including but not limited to angular spectrum propagation, Fresnel diffraction, Fraunhofer diffraction, etc.).

[0095] S5.6: Replace the intensity of the simulated diffraction field above with high dynamic range irradiance, while keeping its phase unchanged, i.e., amplitude constraint in reciprocal space, to obtain the updated simulated diffraction field:

[0096]

[0097] Where || represents taking the element magnitude, F HDR,j This represents the synthesized high dynamic range irradiance at the j-th scan position.

[0098] S5.7: The updated simulated diffraction field is propagated in reverse using a scalar diffraction inverse propagation model to obtain the updated output light:

[0099]

[0100] Among them, prop -1 This is the two-dimensional scalar diffraction inverse propagation model corresponding to step S5.5.

[0101] S5.8: Based on the emitted light before and after amplitude constraint, the update function of the stacked diffraction iterative reconstruction algorithm is applied to simultaneously update the illumination probe and the sample under test. The specific formula used is as follows:

[0102]

[0103]

[0104] Where α and β represent iterative search step size factors, with values ​​between [0,1];

[0105] S5.9: Repeat steps S5.4-S5.8 until all positions have been updated to complete one iteration. Then, use the following formula to calculate the root mean square error of the intensity of the simulated diffraction field and the high dynamic range irradiance at all scan positions:

[0106]

[0107] If the root mean square error is less than the set threshold, it indicates that the iteration has converged, and the complex amplitude information of the illumination probe and the sample to be tested is output; if the root mean square error is greater than the set threshold, the iterative cycle of steps S5.4-S5.9 is repeated.

[0108] like Figure 3 The image shows the amplitude and phase patterns of the test sample and illumination probe used in the simulation process provided in this embodiment. The amplitude information of the test sample used in the simulation experiment is as follows: Figure 3 As shown in Figure (a), the phase information is shown in Figure (b), the amplitude information of the illumination probe is shown in Figure (c), and the phase information is shown in Figure (d). The sample size is 384×384 pixels, and the illumination probe is a Gaussian beam with a wavelength of 632.8 nm and a beam diameter of 64×64 pixels. In order to better simulate the actual scene, the multiple exposure time series is set as [2]. 0 ,2 0.5 ,2 1 ,…,2 8 The precision motion stage moves the sample under test in a two-dimensional plane along a preset scanning path. The preset scanning path has a step size of 14 pixels in the XY direction, scanning a 21×21 area. It simulates detectors with different dynamic ranges to record diffraction field information at each scanning position with different exposure times, and additionally records 20 dark field information. At the same time, low-intensity mixed noise is coupled into the diffraction field information and the dark field information.

[0109] like Figure 4 The image shows the diffraction field obtained by fusing a single-exposure diffraction field and a multi-exposure high dynamic range image during the simulation process provided in this embodiment. (a) is the diffraction field image for a single exposure at 1ms, (b) is the diffraction field image for a single exposure at 16ms, (c) is the diffraction field image for a single exposure at 128ms, and (d) is the diffraction field image obtained by fusing multiple exposure high dynamic range images. The diffraction field obtained through high dynamic range image fusion fully preserves higher-order diffraction information, reduces the influence of mixed noise, and significantly improves the signal-to-noise ratio of the diffraction signal.

[0110] like Figure 5The figures show the single-exposure reconstruction results and the multi-exposure high dynamic range image fusion reconstruction results provided in this embodiment. (a) represents the complex amplitude information of the sample under test reconstructed from a single exposure; (b) represents the complex amplitude information of the sample under test reconstructed from a multi-exposure high dynamic range image fusion reconstruction; (c) represents the complex amplitude information of the illumination probe reconstructed from a single exposure; and (d) represents the complex amplitude information of the illumination probe reconstructed from a multi-exposure high dynamic range image fusion reconstruction.

[0111] like Figure 6 The figure shows the root mean square error curves of single-exposure reconstruction and multi-exposure high dynamic range image fusion reconstruction under different detector dynamic ranges provided in this embodiment. Simulation results show that for the single-exposure reconstruction method, increasing the detector dynamic range can significantly improve the reconstruction quality. For the multi-exposure high dynamic range image fusion reconstruction method, increasing the detector dynamic range is effective when the dynamic range is less than 14 bits, which can improve the image quality. When the dynamic range exceeds 14 bits, the image quality remains basically unchanged. Under the same dynamic range, compared with the single-exposure reconstruction method, the multi-exposure high dynamic range image fusion reconstruction method significantly improves the image reconstruction quality of overlapping correlation imaging. This invention obtains the joint probability based on maximum likelihood estimation and obtains the high dynamic range absolute irradiance by fusing absolute irradiance, which greatly improves the signal-to-noise ratio of the diffraction signal, effectively reduces the requirements for the detector dynamic range, and significantly improves the resolution and contrast of the reconstructed image.

[0112] It is understood that the various numerical designations used in the embodiments of the present invention are merely for the convenience of description and are not intended to limit the scope of the embodiments of the present invention.

[0113] In this article, the term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The symbol " / " in this article indicates that the related objects are in an "or" relationship; for example, A / B means A or B.

[0114] The terms "first" and "second," etc., used in the specification and claims herein are used to distinguish different objects, not to describe a specific order of the objects. For example, "first formula" and "second formula," etc., are used to distinguish different formulas, not to describe a specific order of the formulas.

[0115] In embodiments of the present invention, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in embodiments of the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0116] In the description of the embodiments of the present invention, unless otherwise stated, "multiple" means two or more, for example, multiple scan positions means two or more scan positions, etc.

[0117] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A high dynamic range image fusion and stacked diffraction imaging method, characterized in that, Includes the following steps: S1 initializes the illumination probe and the sample to be tested, acquires the scanning position information, diffraction light field intensity information and dark field information of the sample to be tested at different exposure times, and synthesizes high dynamic range irradiance based on the exposure time, diffraction light field intensity information and dark field information. The S2 simulates the first emitted light generated by the interaction between the illumination probe and the sample at different positions along the scanning path. The first emitted light is then propagated to the detector plane through a two-dimensional scalar diffraction propagation model to obtain a simulated diffraction field. This simulated diffraction field is then used to simulate the detector to measure the first emitted light. S3 replaces the intensity of the simulated diffraction field with high dynamic range irradiance while keeping its phase unchanged to obtain an updated simulated diffraction field; the updated simulated diffraction field is then propagated in reverse using a scalar diffraction inverse propagation model to obtain an updated second outgoing light. S4 Based on the first and second emitted light, the illumination probe and the sample to be tested are updated simultaneously, and the updated illumination probe and the sample to be tested are used as the initial guess for the next position. Steps S2-S3 are repeated until all scanning positions are updated. S5 obtains the root mean square error of the intensity of the simulated diffraction light field and the high dynamic range irradiance at all scanning positions. When the root mean square error is less than the preset threshold, the iterated illumination probe and the sample to be tested are output; otherwise, proceed to step S2. In step S1, the method for synthesizing high dynamic range irradiance includes the following steps: S101 constructs a joint distribution containing diffraction light field information, dark field information, and exposure time series at each scanning position, and obtains the joint probability density function corresponding to the joint distribution; S102 At each scanning position, a high dynamic range absolute irradiance is synthesized based on the joint probability density function; At any scan position, the constructed joint distribution is expressed as the first formula: in, Q i for t i The joint distribution constructed by exposure time, For the process t i Exposure time diffraction field readout; For the process t i Exposure time dark field readout value; This refers to the total gain coefficient during the conversion process; F This refers to the absolute irradiance received by the detector; t i For the first i One exposure time; , , , and The representatives came from , Different independent distributions; U Dark current shot noise, which is a transition from photons to charges. Readout noise for charge-to-digital readout; and It follows a Poisson difference distribution, and approaches a Gaussian distribution when the absolute irradiance is large. It also follows a Gaussian distribution.

2. The high dynamic range image fusion stacked diffraction imaging method as described in claim 1, characterized in that, In the first formula, the conditional probability equals the joint probability of the joint distribution; therefore, based on each exposure time, { Q i } i=1,2,…,N The joint probability density function obtained from conditional probability is expressed as: in, for t i Mathematical expectation of the joint distribution of exposure times; for t i The variance of the joint distribution of exposure times. i= 1,2,...,N ; q i That is Q i ,for t i Joint distribution under exposure time.

3. The high dynamic range image fusion stacked diffraction imaging method as described in claim 1, characterized in that, In step S102, the maximum likelihood estimate of the absolute irradiance is solved according to the joint probability density function, and then the high dynamic range absolute irradiance is synthesized based on the maximum likelihood estimate. The maximum likelihood estimate is obtained through the second formula: in, For the process t i Exposure time diffraction field readout; For the process t i Exposure time dark field readout value; This refers to the total gain coefficient during the conversion process; This represents the absolute irradiance received by the detector; Let be the mathematical expectation of the joint distribution; q i That is Q i ,for t i Joint distribution under exposure time; for t i The variance of the joint distribution of exposure times; This is a mathematical expectation for approximating absolute irradiance.

4. The high dynamic range image fusion stacked diffraction imaging method as described in claim 3, characterized in that, The synthesized high dynamic range absolute irradiance is expressed by the third formula: Among them, the weight function w i Represented as: , F HDR The mathematical expectation of absolute irradiance, which is an approximate estimate of absolute irradiance from diffraction field information and dark field information, is expressed as: 。 5. The high dynamic range image fusion stacked diffraction imaging method as described in claim 1, characterized in that, The updated simulated diffraction field is represented as follows: in, The simulated diffraction field before the update. This indicates taking the element's magnitude. F HDR,j Indicates the first j High dynamic range irradiance at each scan location.

6. The high dynamic range image fusion stacked diffraction imaging method as described in claim 1, characterized in that, In step S4, the fourth formula is used to update both the illumination probe and the sample to be tested simultaneously: in, and This represents different iterative search step size factors, with values ​​between [0,1]. Indicates the first The sample to be tested is in the probe before the scan position is updated. Indicates the first The sample to be tested in the probe after updating at each scanning position; This indicates the illumination probe before the update. This indicates the updated illumination probe; Indicates the first The emitted light at each scan location before the update Indicates the first Updated emitted light at each scan position; * represents real space coordinates; * represents the conjugate operation of a complex matrix; This represents the maximum value of each element in the matrix.

7. The high dynamic range image fusion stacked diffraction imaging method as described in claim 1, characterized in that, In step S6, the formula for calculating the root mean square error of the intensity of the simulated diffraction field and the high dynamic range irradiance at all scanning positions is expressed as follows: in, F HDR,j Indicates the first j High dynamic range irradiance at each scan location The simulated diffraction field before the update in step S5, where MSE is the mean square error.

8. The high dynamic range image fusion stacked diffraction imaging method as described in claim 1, characterized in that, The exposure time series is [2] 0 ,2 0.5 ,2 1 ,…,2 8 ] ms .

9. A stacked diffraction imaging system for implementing the high dynamic range image fusion stacked diffraction imaging method as described in any one of claims 1-8, characterized in that, The stacked diffraction imaging system is either a transmission imaging system or a reflection imaging system.

Images