Image digital compensation method and system for confocal microscope hardware defects

Abstract

The invention provides an image digital compensation method and system for confocal microscope hardware defects, and the method comprises the steps: importing an original observation image, and synchronously receiving a device log; preprocessing the original observation image; according to an equipment log, based on a vector diffraction theory Richardsamp; according to the Wolf model, a point spread function in an ideal state is calculated, and a dynamic non-ideal point spread function is established on the basis of the point spread function in the ideal state; constructing a deconvolution objective function based on the observation image and the dynamic non-ideal point spread function; performing iterative optimization solution on the precondition conjugate gradient for constructing the deconvolution objective function; and outputting a recovery image obtained by iterative optimization solution, and recording the used dynamic non-ideal point spread function and an equipment log. According to the method, high-precision image quality recovery can be realized, the system adaptability and universality are high, the equipment maintenance and use cost can be reduced, the operation process is simplified, the use threshold is reduced, and the method has high efficiency and stability.

Description

Technical Field

[0001] This invention relates to the field of microscopic imaging and image processing / digital optical compensation technology, and in particular to a method and system for digital image compensation of hardware defects in confocal microscopes. Background Technology

[0002] Confocal microscopy, with its excellent optical sectioning capabilities and high spatial resolution, has become a key tool in life sciences, pathological imaging, and materials analysis. However, hardware degradation in actual equipment, such as laser aging leading to increased spot size and detector instability causing a decrease in signal-to-noise ratio, can severely affect image quality and hinder accurate quantitative analysis.

[0003] Currently, traditional solutions to these problems have limitations. They typically rely on expensive and time-consuming hardware replacement or repair, or traditional deconvolution methods that assume an ideal point spread function. The latter cannot effectively compensate for non-ideal point spread functions caused by hardware degradation, especially under low signal-to-noise ratio conditions, where such methods are prone to amplifying noise and producing artifacts. Summary of the Invention

[0004] To achieve the above-mentioned objectives and other advantages of the present invention, a first objective of the present invention is to provide a method for digital image compensation of hardware defects in a confocal microscope, comprising the following steps: Import the original observation images and receive the device logs synchronously; The original observed images are preprocessed; Based on the device logs, and using the Richards & Wolf model of vector diffraction theory, the point spread function under ideal conditions is calculated, and a dynamic non-ideal point spread function is established based on the point spread function under ideal conditions. Construct a deconvolution objective function based on the observed image and the dynamic non-ideal point spread function; The preconditional conjugate gradients for constructing the deconvolution objective function are iteratively optimized and solved. Output the restored image obtained by iterative optimization, and record the dynamic non-ideal point spread function and device log used.

[0005] Furthermore, the device log includes configuration parameters, including objective lens numerical aperture, excitation wavelength, emission wavelength, medium refractive index, pixel size, and pinhole size.

[0006] Furthermore, the preprocessing step for the original observed image includes: Remove background offset; Adjust pixel values ​​that exceed the preset grayscale range.

[0007] Furthermore, the step of calculating the point spread function under ideal conditions based on the device logs and the Richards & Wolf model of vector diffraction theory includes: Establish a parameter model of the optical imaging system based on the configuration parameters; The focused electric field is decomposed using vector diffraction theory to obtain a complete three-dimensional electric field distribution within the focal region; The intensity superposition operation is performed on the three-dimensional electric field distribution to obtain the point diffusion intensity distribution in three-dimensional space; The point diffusion intensity distribution is normalized to obtain the theoretical point diffusion function. .

[0008] Furthermore, the step of establishing a dynamic non-ideal point spread function based on the point spread function under the ideal state includes: Sparsely distributed samples were prepared using nanospheres with a diameter no larger than a preset value and bright fluorescence. Multiple microspheres were acquired using a target confocal microscope with configuration parameters consistent with actual imaging in a 3DZ-stack. For each microsphere, the maximum intensity projection was performed, and the full width at half maximum (FWHM) of the point spread function was estimated using three-dimensional Gaussian fitting, yielding the transverse values ​​respectively. and axial Numerical value; The broadening coefficient is obtained by comparing the full width at half maximum (FWHM) of the diffusion function at the theoretical point with that of the theoretical point. , = ; Use the expansion factor , Construct a Gaussian broadening kernel to correct the theoretical point spread function, and the standard deviation of the broadening kernel is... , Determined by the broadening factor: The final corrected expression for the point spread function is: in, This represents the convolution operation. A three-dimensional Gaussian broadened kernel.

[0009] Furthermore, the step of establishing a dynamic non-ideal point spread function based on the point spread function under the ideal state includes: Define a point spread function to broaden the search space of the parameters; Traverse each candidate parameter combination in the search space; wherein, for each combination, perform the following operations: The theoretical point spread function is convolved with a three-dimensional Gaussian broadening kernel defined by the candidate parameter combination to obtain the candidate point spread function. The candidate point diffusion function is used to perform a fast deconvolution algorithm based on Wiener inverse filtering on the original observed image to obtain preliminary restoration results; Calculate the high-frequency energy of the spectrum of the restored image; After iterating through all candidate point spread functions, the point spread function that maximizes the sharpness index is selected and used as the adaptive point spread function for the current image.

[0010] Furthermore, the deconvolution objective function is: in, To observe the image, For dynamic non-ideal point spread function, For the image to be recovered, Represents spatial convolution. For data fidelity items, For regularization terms, This represents the regularization strength.

[0011] Furthermore, the step of constructing the deconvolution objective function based on the observed image and the dynamic non-ideal point spread function includes: For each pixel in the image, calculate the local mean and local standard deviation of that pixel within a window around it to estimate the local SNR. By taking the local SNR as input and substituting it into the nonlinear mapping function, the regularization parameter corresponding to that pixel is calculated. ; Generate a spatially adaptive regularization parameter graph This is then applied to the subsequent optimization solution process.

[0012] Furthermore, the step of iteratively optimizing the preconditional conjugate gradient of the deconvolution objective function includes: The image to be estimated is initialized using the pixel mean of the input observed image Y. ; Y, The frequency domain is transformed by Fourier transform. Calculate the preconditioner ; Calculate the initial gradient Calculate the initial preprocessing gradient ; Set the initial search direction Initial iteration count Maximum number of iterations ; Calculate step size Among them, step size The goal is to minimize along the current search direction objective function By approximating the objective function, the optimal step size is: ; Update image = ; Forced nonnegativity constraint This operation is performed in the airspace. The result is converted to the spatial domain, and then Fourier transformed to the frequency domain to continue the iteration. Update gradient Update the preprocessed gradient ; Calculate conjugate parameters ; Update search direction ; like Or reach the maximum number of iterations. If the iteration stops, return the current iteration result. ; Update iteration count .

[0013] A second objective of this invention is to provide an image digital compensation system for hardware defects in a confocal microscope, employing the aforementioned method, comprising a data acquisition unit, a preprocessing unit, a point spread function modeling unit, a deconvolution reconstruction unit, and an output unit; wherein, The data acquisition unit is used to import the original observation images and simultaneously receive the device logs; The preprocessing unit is used to preprocess the original observation image; The point spread function modeling unit is used to calculate the point spread function under ideal conditions based on the device log and the Richards & Wolf model of vector diffraction theory, and to establish a dynamic non-ideal point spread function based on the point spread function under ideal conditions. The deconvolution reconstruction unit is used to construct a deconvolution objective function based on the observed image and the dynamic non-ideal point spread function, and to iteratively optimize and solve the preconditional conjugate gradient for constructing the deconvolution objective function. The output unit is used to output the restored image obtained by iterative optimization and to record the dynamic non-ideal point spread function and device log used.

[0014] A third objective of the present invention is to provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method.

[0015] A fourth objective of the present invention is to provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described method.

[0016] Compared with the prior art, the beneficial effects of the present invention are: High-precision image quality restoration: This invention is significantly superior to traditional deconvolution methods in terms of resolution improvement and noise suppression, achieving imaging results close to ideal hardware conditions.

[0017] Specifically, relying on dynamic adaptive PSF modeling, this invention can establish a non-ideal PSF that matches the actual situation for hardware defects, rather than relying on an ideal model, thereby achieving more accurate fuzzy compensation. This mechanism is implemented through fluorescent microsphere calibration (Implementation A) or data-driven adaptation (Implementation B), ensuring that the recovery results are more consistent with the actual hardware characteristics.

[0018] Adaptive regularization is introduced in the deconvolution optimization target construction step, which dynamically adjusts the regularization intensity according to the local signal-to-noise ratio of the image: regularization is reduced in areas with rich details to preserve information, and regularization is enhanced in areas with dominant noise to effectively suppress noise. This achieves a balance between resolution improvement and noise suppression, avoiding the drawback of traditional methods that tend to amplify noise under low signal-to-noise ratio conditions.

[0019] System adaptability and universality: This invention has good versatility, does not depend on specific hardware configurations, and can adapt to various confocal microscope systems and different degrees of hardware degradation.

[0020] Among them, dynamic adaptive PSF modeling is the core, especially implementation method B, which determines the optimal PSF by analyzing the characteristics of the image to be processed. It can run without any calibration experiments, thereby achieving universal compatibility with microscopes of different models and aging conditions.

[0021] Configuration parameter synchronization and PSF traceability records provide complete data support for system adaptive operation and subsequent maintenance, further improving cross-system stability and reliability.

[0022] Cost-effectiveness and user-friendliness: This invention can reduce equipment maintenance and usage costs, while simplifying the operation process and lowering the barrier to entry.

[0023] Specifically, through software compensation strategies, users can obtain high-quality images without expensive hardware replacements or complex optical path adjustments, greatly reducing maintenance costs.

[0024] The fully automated adaptive optimization process in Implementation Method B allows users to directly obtain optimized recovery results without requiring specialized knowledge or additional experiments (such as microsphere calibration). This operating mode greatly enhances user experience and ease of use.

[0025] High efficiency and stability: While ensuring image quality, this invention can achieve fast and stable calculations, making it suitable for processing large-scale 3D imaging data.

[0026] The preconditional conjugate gradient method used in the iterative solution and constraint steps has excellent convergence and computational efficiency. Furthermore, in the image and variable initialization steps, the convolution operation is transferred to the frequency domain through Fourier transform, which further accelerates the iterative solution process.

[0027] The mandatory nonnegativity constraint introduced in the iterative loop step ensures that the pixel values ​​of the recovered image conform to physical reality, avoids the generation of artifacts, and thus improves the physical rationality of the results and the stability of the algorithm.

[0028] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it according to the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Specific embodiments of the present invention are given in detail below with reference to the accompanying drawings. Attached Figure Description

[0029] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 This is a schematic diagram illustrating the principle of digital image compensation for hardware defects in a confocal microscope. Figure 2 This is a schematic diagram of a computer device. Figure 3 This is a schematic diagram of a computer-readable storage medium. Detailed Implementation

[0030] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. It should be noted that, without conflict, the various embodiments or technical features described below can be arbitrarily combined to form new embodiments.

[0031] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention.

[0032] The drawing numbers in this application are only used to distinguish the steps in the scheme and are not used to limit the execution order of the steps. The specific execution order is as described in the specification.

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0034] To address the aforementioned technical problems, this invention proposes an innovative software compensation method: combining fluorescent microsphere calibration with data-driven adaptive point spread function modeling, and integrating it into a regularized deconvolution computer implementation process. This method enables precise digital compensation for hardware defects without requiring hardware replacement, significantly improving image resolution and signal-to-noise ratio. This solution boasts high feasibility, stability, and versatility, meeting the needs of research and industry for stable and traceable image quality improvement, and providing an economical, efficient, and reliable solution. The specific solution is as follows: Example 1

[0035] A method for digital image compensation of hardware defects in confocal microscopes, such as Figure 1 As shown, it includes the following steps: S1. Import the original observation images and receive the device logs simultaneously; In some embodiments, raw observation images are imported from a confocal microscope acquisition module or storage medium. .

[0036] Furthermore, the device log includes the objective lens numerical aperture. Excitation wavelength , emitted light wavelength medium refractive index Configuration parameters such as pixel size and pinhole size.

[0037] S2. Preprocess the original observation image; Furthermore, the preprocessing step for the original observed image includes: Background subtraction, removing background offset; Pixel range cropping automatically adjusts pixel values ​​that exceed a preset grayscale range. Specifically, pixels above the range are set to the highest grayscale value, and pixels below the range are set to the lowest grayscale value.

[0038] S3. Based on the device log, and using the Richards & Wolf model of vector diffraction theory, calculate the point spread function (PSF) under ideal conditions, and establish a dynamic non-ideal point spread function based on the point spread function under ideal conditions. In some embodiments, the step of basing decisions on the device logs and using vector diffraction theory by Richards & The Wolf model, in calculating the point spread function under ideal conditions, includes the following steps: Establish a parameter model of the optical imaging system based on the configuration parameters; The focused electric field is decomposed using vector diffraction theory to obtain a complete three-dimensional electric field distribution within the focal region; The intensity superposition operation is performed on the three-dimensional electric field distribution to obtain the point diffusion intensity distribution in three-dimensional space; The point diffusion intensity distribution is further normalized to obtain the theoretical point diffusion function. .

[0039] This invention obtains the theoretical point diffusion function Based on this, a dynamic non-ideal point spread function is established. Two implementation methods are provided, which can be used in parallel or selectively.

[0040] In some embodiments, a fluorescent microsphere calibration implementation method (Implementation Method A) is used. Specifically, the step of establishing a dynamic non-ideal point diffusion function based on the point diffusion function under the ideal state includes: A1. Sample preparation: Select nanospheres with a diameter not greater than a preset value (e.g., a diameter not greater than 50 nm) and bright fluorescence to prepare sparsely distributed samples; A2. Imaging Acquisition: Using a target confocal microscope with configuration parameters consistent with actual imaging, acquire 3DZ-stack images of multiple microspheres; A3. Data Processing: For each microsphere, maximum intensity projection was performed, and the full width at half maximum (FWHM) of the point spread function was estimated using three-dimensional Gaussian fitting. The transverse values ​​were then obtained. and axial Numerical value; A4. Theoretical Comparison: Compare the full width at half maximum (FWHM) of the diffusion function at the theoretical point to obtain the broadening coefficient. , = ; A5, PSF Correction: Using the widening factor , Construct a Gaussian broadening kernel to correct the theoretical point spread function, and the standard deviation of the broadening kernel is... , Determined by the broadening factor: The final corrected expression for the point spread function is: in, This represents the convolution operation. A three-dimensional Gaussian broadened kernel.

[0041] In other embodiments, a data-driven adaptive implementation method based on the image to be processed (Implementation Method B) is adopted. This eliminates the need for prior fluorescent microsphere calibration; the PSF can be adaptively determined simply by analyzing the features of the image itself, thereby compensating for blurring caused by hardware degradation. Specifically, the step of establishing a dynamic non-ideal point spread function based on the point spread function under the ideal state includes: B1. Define the point spread function to broaden the search space of the parameters; B2. Optimization Search and Evaluation of PSF Parameters: The computer traverses each candidate parameter combination within the parameter search space defined above. For each combination, the computer performs the following operations: B2.1 Generating a candidate PSF: Perform a three-dimensional convolution between the theoretical point spread function and a three-dimensional Gaussian broadening kernel defined by the candidate parameter combination to obtain a candidate point spread function; B2.2 Perform fast deconvolution: Use the candidate point spread function to perform a fast deconvolution algorithm based on Wiener deconvolution on the original observed image to obtain preliminary restoration results; B2.3 Calculation of Sharpness Index: This index calculates the high-frequency energy of the restored image's spectrum. It is achieved by performing a Fourier transform on the restored image and calculating the sum of the energy in the high-frequency components of its spectrum. High-frequency energy accurately reflects the degree of detail and edge restoration in the image; a higher value indicates a sharper image.

[0042] B2.4 Determining the Optimal PSF: After traversing all candidate point spread functions, the computer selects the point spread function that maximizes the sharpness index and uses it as the adaptive point spread function for the current image. .

[0043] S4. Construct a deconvolution objective function based on the observed image and the dynamic non-ideal point spread function; Furthermore, the deconvolution objective function is: in, To observe the image, For dynamic non-ideal point spread function, For the image to be recovered, Represents spatial convolution. For data fidelity items, For regularization terms, This represents the regularization strength.

[0044] Known quantities in the deconvolution objective function: observed image Y, dynamic non-ideal point spread function H; Unknown quantity: Image to be recovered ; Data fidelity item Poisson noise model estimation for matched confocal imaging uses the negative log-likelihood approximation. ; Regular terms : Used for noise suppression and edge preservation. To prevent smoothing terms with zero gradient; Regularization strength : Dynamically adjust based on local features of the image.

[0045] Furthermore, the step of constructing the deconvolution objective function based on the observed image and the dynamic non-ideal point spread function includes: Estimating the local signal-to-noise ratio (SNR): For each pixel in an image, the local mean and local standard deviation of that pixel are calculated within a window surrounding it to estimate the local SNR. This local SNR reflects the signal and noise levels of the image at the current location. Calculate the adaptive regularization parameter: Take the local SNR as input, substitute it into the nonlinear mapping function, and calculate the regularization parameter corresponding to that pixel. In the high SNR region, A smaller value reduces the regularization intensity, preserving image details to the maximum extent; in low SNR regions, A larger value enhances the regularization strength and effectively suppresses noise; Generate a spatially adaptive regularization parameter graph This is then applied to the subsequent optimization solution process.

[0046] S5. Iteratively optimize and solve the preconditional conjugate gradient for constructing the deconvolution objective function; Furthermore, the step of iteratively optimizing the preconditional conjugate gradient of the deconvolution objective function includes: The image to be estimated is initialized using the pixel mean of the input observed image Y. ; Y, The frequency domain is transformed by Fourier transform. Calculate the preconditioner ; Calculate the initial gradient Calculate the initial preprocessing gradient ; Set the initial search direction Initial iteration count Maximum number of iterations ; Calculate step size Step size The goal is to minimize along the current search direction objective function By approximating the objective function, the optimal step size is... .

[0047] Update image : = ; Mandatory nonnegativity constraint: This operation is performed in the airspace. The result is converted to the spatial domain, and then Fourier transformed to the frequency domain to continue the iteration. Update gradient Update the preprocessed gradient ; Calculate conjugate parameters : ; Update search direction : ; Check for convergence: If Or reach the maximum number of iterations. If the iteration stops, return the current iteration result. ; Number of updates / iterations: .

[0048] S5. Output the restored image obtained by iterative optimization, and record the dynamic non-ideal point spread function and device log used.

[0049] For example, output the restored image obtained in step S5. To the device or storage medium: record the dynamic non-ideal point spread function used. With configuration parameters, so that it can be traced.

[0050] This invention provides an image processing method executed by a computer or other device with data processing capabilities for digitally compensating for degraded images from confocal microscopes caused by hardware defects, such as laser aging or uneven detector gain. This method outputs high-resolution, low-noise restored images through a joint framework based on dynamic non-ideal point spread function modeling and regularized deconvolution. Example 2

[0051] A digital image compensation system for hardware defects in a confocal microscope is provided, employing the method described above. For a detailed description of the method, please refer to the corresponding description in the above method embodiments; it will not be repeated here. Figure 1 As shown, the system includes a data acquisition unit, a preprocessing unit, a point spread function modeling unit, a deconvolution reconstruction unit, and an output unit; among which, The data acquisition unit is used to import the original observation images and simultaneously receive the device logs; The preprocessing unit is used to preprocess the original observation image; The point spread function modeling unit is used to calculate the point spread function under ideal conditions based on the device log and the Richards & Wolf model of vector diffraction theory, and to establish a dynamic non-ideal point spread function based on the point spread function under ideal conditions. The deconvolution reconstruction unit is used to construct a deconvolution objective function based on the observed image and the dynamic non-ideal point spread function, and to iteratively optimize and solve the preconditional conjugate gradient for constructing the deconvolution objective function. The output unit is used to output the restored image obtained by iterative optimization and to record the dynamic non-ideal point spread function and device log used. Example 3

[0052] A computer device 100, such as Figure 2 As shown, the system includes a memory 110, a processor 120, and a computer program 130 stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of an image digital compensation method for hardware defects in a confocal microscope. For a detailed description of the method, please refer to the corresponding description in the above method embodiments; it will not be repeated here. Example 4

[0053] A computer-readable storage medium, such as Figure 3 As shown, a computer program is stored thereon. When executed by a processor, the computer program implements the steps of an image digital compensation method for hardware defects in a confocal microscope. For a detailed description of the method, please refer to the corresponding description in the above method embodiments, which will not be repeated here.

[0054] The number of devices and processing scale described herein are for the purpose of simplifying the description of the invention. Applications, modifications, and variations of the invention will be readily apparent to those skilled in the art.

[0055] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

[0056] The apparatus, computer device, and non-volatile computer storage medium and method provided in the embodiments of this specification are corresponding. Therefore, the apparatus, computer device, and non-volatile computer storage medium also have similar beneficial technical effects as the corresponding method. Since the beneficial technical effects of the method have been described in detail above, the beneficial technical effects of the corresponding apparatus, computer device, and non-volatile computer storage medium will not be repeated here.

[0057] Those skilled in the art will also know that, besides implementing the controller in the form of purely computer-readable program code, the same functions can be achieved by logically programming the method steps, making the controller take the form of logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers (PLCs), and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the devices included within it for implementing various functions can also be considered structures within that hardware component. Alternatively, the devices for implementing various functions can be considered as both software units implementing the method and structures within a hardware component.

[0058] The systems, apparatuses, or units described in the above embodiments can be implemented by computer chips or physical entities, or by products with certain functions. For ease of description, the above apparatuses are described separately as various units based on their functions. Of course, when implementing one or more embodiments of this specification, the functions of each unit can be implemented in one or more software and / or hardware.

[0059] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the embodiments of this specification can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0060] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0061] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0062] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0063] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0064] This specification may be described in the general context of computer-executable instructions, such as program units, that are executed by a computer. Generally, program units include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification may also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program units may reside in local and remote computer storage media, including storage devices.

[0065] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0066] The above description is merely an embodiment of this specification and is not intended to limit the scope of one or more embodiments of this specification. Various modifications and variations can be made to one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of one or more embodiments of this specification.

Claims

1. An image digitalization compensation method for confocal microscopy hardware defects, characterized in that, The method comprises the following steps: Importing an original observation image and synchronously receiving a device log; Preprocessing the original observation image; According to the device log, calculating a point spread function in an ideal state based on a Richards & Wolf model of vector diffraction theory, and establishing a dynamic non-ideal point spread function based on the point spread function in the ideal state; Constructing a deconvolution target function based on the observation image and the dynamic non-ideal point spread function; Iteratively optimizing and solving precondition conjugate gradients of the constructed deconvolution target function; Outputting a restored image obtained by the iterative optimization and solution, and recording the dynamic non-ideal point spread function and the device log.

2. A method for image digitalization compensation of confocal microscopy hardware defects as claimed in claim 1, characterized in that: The device log comprises configuration parameters, and the configuration parameters comprise an objective numerical aperture, an excitation light wavelength, an emission light wavelength, a medium refractive index, a pixel size and a pinhole size.

3. A method for image digitalization compensation of confocal microscopy hardware defects as claimed in claim 1, wherein, The preprocessing step of the original observation image comprises: Removing a background bias; Adjusting pixel values beyond a preset gray scale range.

4. A method for image digitalization compensation of confocal microscopy hardware defects as claimed in claim 2, wherein, The step of calculating a point spread function in an ideal state based on a Richards & Wolf model of vector diffraction theory according to the device log comprises: Establishing a parameter model of an optical imaging system according to the configuration parameters; Decomposing a focused electric field to obtain a complete three-dimensional electric field distribution in a focal point region through vector diffraction theory; Performing intensity superposition operation on the three-dimensional electric field distribution to obtain a point spread intensity distribution in a three-dimensional space; normalizing the point spread intensity distribution to obtain a theoretical point spread function number.

5. A method for image digitalization compensation of confocal microscopy hardware defects as claimed in claim 4, wherein, The step of establishing a dynamic non-ideal point spread function based on the point spread function in the ideal state comprises: Selecting nanometer microspheres with a diameter not greater than a preset value and bright fluorescence to prepare a sparse distribution sample; Collecting 3D Z-stacks of a plurality of microspheres on a target confocal microscope with consistent configuration parameters as actual imaging; The maximum intensity projection was performed for each microsphere, and the full width at half maximum of the point spread function was estimated by three-dimensional Gaussian fitting. The lateral and axial numerical values were obtained, respectively. a full width at half maximum corresponding to the theoretical point spread function is compared to obtain a broadening coefficient , ; Using a spreading factor , Constructing a Gaussian spreading kernel to modify the theoretical point spread function, the spreading kernel standard deviation , is determined from the spreading factor: An expression of the finally corrected point spread function is: wherein denotes a convolution operation, is a three-dimensional Gaussian spread kernel.

6. A method for image digitalization compensation of hardware defects for a confocal microscope as claimed in claim 4 or 5, characterized in that, The step of establishing a dynamic non-ideal point spread function based on the point spread function in the ideal state comprises: Defining a search space of point spread function widening parameters; Iterating each candidate parameter combination in the search space; wherein, for each combination, the following operations are performed: Performing three-dimensional convolution on the theoretical point spread function and a three-dimensional Gaussian widening kernel defined by the candidate parameter combination to obtain a candidate point spread function; Using the candidate point spread function to perform a fast deconvolution algorithm based on Wiener inverse filtering on the original observation image to obtain a preliminary restoration result; Calculating a high-frequency energy of a spectrum of the restored image; After iterating all candidate point spread functions, selecting a point spread function that makes a sharpness index reach a maximum value from the candidate point spread functions as an adaptive point spread function of the current image.

7. A method for image digitalization compensation of confocal microscopy hardware defects as claimed in claim 1 wherein, The deconvolution target function is: wherein is an observed image, is a dynamic non-ideal point spread function, is an image to be restored, denotes a spatial convolution, is a data fidelity term, is a regularization term, is a regularization strength.

8. A method for image digitalization compensation of confocal microscopy hardware defects as claimed in claim 7, wherein, The step of constructing a deconvolution target function based on the observation image and the dynamic non-ideal point spread function comprises: For each pixel point in the image, calculating a local mean and a local standard deviation of the pixel in a window around the pixel point to estimate a local SNR; The local SNR is taken as an input, and is brought into a nonlinear mapping function to calculate the regularization parameter corresponding to the pixel point ; Generating spatially adaptive regularization parameter maps and applying it to a subsequent optimization solution process.

9. A method for image digitalization compensation of confocal microscopy hardware defects as claimed in claim 1 wherein, The step of iteratively optimizing and solving precondition conjugate gradients of the constructed deconvolution target function comprises: Using input observation images of the input observation images ; The , are Fourier transformed from the spatial domain to the frequency domain; Computing a preconditioner ; computing an initial gradient , computing an initial pre-processed gradient ; Setting initial search direction , initial iteration count , maximum iteration count ; Computing step size ; where step size The goal is to minimize the objective function along the current search direction By approximating the objective function, the optimal step size is ; updating the image ; Enforcing non-negativity constraints This operation is done in the spatial domain, converting to the spatial domain, and then Fourier transforming the result to the frequency domain to continue the iteration; updating gradient , updating pre-processed gradient ; Computing conjugate parameters ; updating the search direction ; If , or the maximum iteration number is reached , the iteration is stopped and the current iteration result is returned ; number of update iterations .

10. An image digitization compensation system for confocal microscopy hardware defects, applying the method of any one of claims 1 to 9, characterized in that: The device comprises a data acquisition unit, a preprocessing unit, a point spread function modeling unit, a deconvolution reconstruction unit and an output unit. The data acquisition unit is configured to import a raw observation image and synchronously receive a device log; The preprocessing unit is configured to preprocess the raw observation image; The point spread function modeling unit is configured to calculate a point spread function in an ideal state based on a vector diffraction theory Richards & Wolf model according to the device log, and establish a dynamic non-ideal point spread function based on the point spread function in the ideal state; The deconvolution reconstruction unit is configured to construct a deconvolution objective function based on the observation image and the dynamic non-ideal point spread function, and iteratively optimize and solve a preconditioned conjugate gradient of the constructed deconvolution objective function; The output unit is configured to output a recovered image obtained by the iterative optimization and solving, and record the used dynamic non-ideal point spread function and the device log.

11. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1-9.

12. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method according to any one of claims 1-9.

Images