A method and apparatus for super-resolution image scanning microscopy reconstruction based on frequency decomposition

Abstract

This invention relates to the fields of optical imaging and image processing technology, and discloses an image scanning microscopy super-resolution reconstruction method and apparatus based on frequency decomposition. The method includes: acquiring a set of original image data blocks; wherein the set of original image data blocks includes multiple original dot matrix images obtained by two-dimensional scanning of a target sample using an image scanning microscopy system; performing multi-level frequency decomposition on the original dot matrix images to filter out the highest frequency noise component, obtaining multiple two-dimensional frequency components; fusing the multiple two-dimensional frequency components to generate a binarized digital pinhole mask; and extracting and synthesizing high-frequency information based on the binarized digital pinhole mask and the highest two-dimensional frequency component among the multiple two-dimensional frequency components to obtain a super-resolution reconstructed image. This invention provides a faster, more stable, and simpler super-resolution imaging tool.

Description

Technical Field

[0001] This invention relates to the field of optical imaging and image processing technology, and specifically to an image scanning microscopic super-resolution reconstruction method and apparatus based on frequency decomposition. Background Technology

[0002] Image scanning microscopy is a novel super-resolution fluorescence microscopy technique that not only breaks the optical diffraction limit on imaging resolution but also possesses strong tissue penetration and tomographic imaging capabilities.

[0003] Currently, super-resolution reconstruction methods based on image scanning microscopy employ pixel redistribution algorithms. However, these algorithms suffer from problems such as slow processing speed, sensitivity to signal-to-noise ratio, complex processes, and insufficient information utilization, leading to reduced super-resolution reconstruction quality and limiting further improvements in resolution. Summary of the Invention

[0004] This invention provides a method and apparatus for super-resolution image scanning microscopy reconstruction based on frequency decomposition, in order to solve the problems of slow processing speed, sensitivity to signal-to-noise ratio and complex process of related super-resolution reconstruction methods.

[0005] In a first aspect, the present invention provides an image scanning microscopic super-resolution reconstruction method based on frequency decomposition, the method comprising:

[0006] Obtain a set of raw image data blocks; wherein, the set of raw image data blocks includes multiple raw raster images obtained by two-dimensional scanning of the target sample using an image scanning microscopy system;

[0007] The original raster image is subjected to multi-level frequency decomposition to filter out the highest frequency noise component, resulting in multiple two-dimensional frequency components.

[0008] Multiple two-dimensional frequency components are fused to generate a binary digital pinhole mask.

[0009] Based on a binarized digital pinhole mask and the highest two-dimensional frequency component among multiple two-dimensional frequency components, high-frequency information is extracted and synthesized to obtain a super-resolution reconstructed image.

[0010] This invention provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition. By performing multi-level frequency decomposition on the original dot matrix image, the original dot matrix image is decomposed into multiple two-dimensional frequency components. These multiple two-dimensional frequency components are then fused to obtain a binarized digital pinhole mask. The standardized process of frequency decomposition-fusion-masking significantly reduces the professional requirements for operators, improves the repeatability and ease of use of the method, and expands its application scenarios. Furthermore, based on the binarized digital pinhole mask and the highest two-dimensional frequency component among the multiple two-dimensional frequency components, the complete high-frequency component after adaptive masking is directly extracted and retained, making fuller use of the photon information collected by the detector. This theoretically possesses the potential to achieve even higher resolution limits. Through the technical path of frequency separation and parallel processing, direct super-resolution image synthesis without sub-pixel localization, digital pinholes, or Fourier transforms is achieved, improving the processing speed of super-resolution imaging and resulting in super-resolution reconstructed images with high signal-to-noise ratio and high resolution.

[0011] In one optional implementation, the original bitmap image undergoes multi-level frequency decomposition to filter out the highest frequency noise component, resulting in multiple two-dimensional frequency components, including:

[0012] The original dot matrix image is subjected to frequency sieving to obtain candidate mode functions;

[0013] Subtract the candidate mode function from the original bitmap image to obtain the residual components;

[0014] The residual components are iteratively decomposed to obtain multiple mode functions;

[0015] Multiple modal functions are sorted from high to low frequency to obtain multiple modal components;

[0016] The highest frequency noise component is extracted from multiple modal components and then filtered out to obtain multiple two-dimensional frequency components.

[0017] This invention provides an image scanning microscopic super-resolution reconstruction method based on frequency decomposition. It iteratively separates the modal components from high to low frequencies in the original bitmap image, ultimately obtaining modal functions and residual components at different frequencies. Following a separation logic of high frequencies first, then low frequencies, the first obtained modal component corresponds to the highest frequency information of the image, followed by components corresponding to mid-frequency texture and low-frequency background. The final residual component represents the stable trend base of the image, with clear layer boundaries. This method ensures the orderliness of frequency decomposition while maximizing the preservation of local detail features of the original bitmap image, providing high-quality and clearly layered foundational data for subsequent component processing and super-resolution reconstruction.

[0018] In one optional implementation, frequency screening is performed on the original bitmap image to obtain candidate mode functions, including:

[0019] The original raster image is initialized to obtain local extreme points;

[0020] Surface interpolation is performed on local extreme points to obtain the upper and lower envelope surfaces, and the mean envelope surface is calculated based on the upper and lower envelope surfaces.

[0021] The difference between the original dot matrix image and the mean envelope is calculated to obtain the candidate mode functions;

[0022] The candidate mode function is compared with the mode function condition. If the candidate mode function satisfies the mode function condition, the candidate mode function is output.

[0023] The present invention provides an image scanning microscopic super-resolution reconstruction method based on frequency decomposition, which does not require complex Fourier transform and is completed directly through a specific envelope function, thus preserving the effective mid-to-high frequency structural information in the image block.

[0024] In one optional implementation, the highest frequency noise component is extracted from multiple modal components, and then filtered out to obtain multiple two-dimensional frequency components, including:

[0025] Calculate the structural similarity index among multiple modal components, and determine the highest frequency noise component based on the structural similarity index;

[0026] The highest frequency noise component among multiple modal components is separated and filtered out to obtain multiple two-dimensional frequency components.

[0027] The present invention provides an image scanning microscopic super-resolution reconstruction method based on frequency decomposition, which accurately distinguishes noise components from effective components by quantifying the correlation of modal component features through structural similarity index, accurately separates the highest frequency noise component, and maximizes the preservation of effective two-dimensional frequencies.

[0028] In one optional implementation, multiple two-dimensional frequency components are fused to generate a binarized digital pinhole mask, including:

[0029] Multiple two-dimensional frequency components are fused together to obtain an intermediate image;

[0030] Based on the intermediate image, a solid circular structure is identified and extracted using an automatic detection algorithm for solid circular spots; the solid circular structure conforms to the characteristics of diffraction-limited light spots.

[0031] The pixel grayscale information corresponding to the solid circular structure is binarized to obtain a binarized digital pinhole mask.

[0032] This invention provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition. It fuses multiple two-dimensional frequency components to obtain an intermediate image, preserving the morphological features of the light spot, improving the signal-to-noise ratio, and effectively suppressing background noise. Furthermore, it utilizes an automatic detection algorithm for solid circular spots to identify and extract solid circular structures, accurately identifying diffraction-limited light spots, reducing manual intervention, and improving processing speed. Finally, it binarizes the pixel grayscale information corresponding to the solid circular structures, converting them into a binary digital pinhole mask. For light spots under different samples and imaging conditions, the mask can automatically adjust the number and distribution of pinholes without manual parameter modification. The binary digital pinhole mask dynamically adapts to the light spot features, supporting super-resolution reconstruction and solving the problems of poor image quality, difficult feature localization, and poor adaptability to subsequent processing.

[0033] In one optional implementation, high-frequency information is extracted and synthesized based on a binarized digital pinhole mask and the highest two-dimensional frequency component among multiple two-dimensional frequency components to obtain a super-resolution reconstructed image, including:

[0034] The binarized digital pinhole mask and the highest two-dimensional frequency component are multiplied pixel by pixel to obtain a super-resolution image patch;

[0035] The pixels corresponding to the super-resolution image patches are mapped in parallel to the global grid to generate the initial super-resolution image;

[0036] Post-processing is performed on the initial super-resolution image to obtain the super-resolution reconstructed image.

[0037] The present invention provides an image scanning microscopic super-resolution reconstruction method based on frequency decomposition, which makes fuller use of the photon information collected by the detector by directly extracting and retaining the complete high-frequency components after adaptive masking, thus theoretically having the potential to explore higher limit resolution.

[0038] In a second aspect, the present invention provides an image scanning microscopic super-resolution reconstruction apparatus based on frequency decomposition, the apparatus comprising:

[0039] The acquisition module is used to acquire a set of raw image data blocks; wherein, the set of raw image data blocks includes multiple raw dot matrix images obtained by performing two-dimensional scanning of the target sample using an image scanning microscopy system;

[0040] The multi-level frequency decomposition module is used to perform multi-level frequency decomposition on the original dot matrix image, filter out the highest frequency noise component, and obtain multiple two-dimensional frequency components.

[0041] The fusion processing module is used to fuse multiple two-dimensional frequency components to generate a binary digital pinhole mask.

[0042] The extraction and synthesis module is used to extract and synthesize high-frequency information based on a binarized digital pinhole mask and the highest two-dimensional frequency component among multiple two-dimensional frequency components to obtain a super-resolution reconstructed image.

[0043] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the image scanning microscopic super-resolution reconstruction method based on frequency decomposition described in the first aspect or any corresponding embodiment thereof.

[0044] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the image scanning microscopic super-resolution reconstruction method based on frequency decomposition described in the first aspect or any corresponding embodiment thereof.

[0045] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the frequency decomposition-based image scanning microscopic super-resolution reconstruction method of the first aspect or any corresponding embodiment described above. Attached Figure Description

[0046] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0047] Figure 1 This is a schematic diagram of an application scenario according to an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of the first process of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition according to an embodiment of the present invention;

[0049] Figure 3 This is a schematic diagram of the second process of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition according to an embodiment of the present invention;

[0050] Figure 4 This is a schematic diagram of the third process of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition according to an embodiment of the present invention;

[0051] Figure 5 This is a schematic diagram of the fourth process of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition according to an embodiment of the present invention;

[0052] Figure 6 This is a flowchart illustrating a super-resolution imaging algorithm based on Fourier transform and pixel redistribution according to an embodiment of the present invention.

[0053] Figure 7 This is a schematic diagram of the overall process of the image scanning microscopic super-resolution reconstruction method based on frequency decomposition according to an embodiment of the present invention;

[0054] Figure 8 This is a flowchart illustrating the frequency decomposition algorithm according to an embodiment of the present invention;

[0055] Figure 9 This is a structural block diagram of an image scanning microscopic super-resolution reconstruction device based on frequency decomposition according to an embodiment of the present invention;

[0056] Figure 10 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0059] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0060] As an optional application scenario of this invention, such as Figure 1 As shown, an image scanning microscopy super-resolution reconstruction system based on frequency decomposition may include at least one terminal device and at least one server. Figure 1 The system is illustrated in the example, which includes a computer 101, a mobile terminal 102, and a server 103, and the terminal devices such as the computer 101 and the mobile terminal 102 are connected to the server 103 through a network 110.

[0061] Specifically, the terminal device can be a smartphone, tablet, laptop, PDA, desktop computer, game console, smart TV, smart wearable device, in-vehicle terminal, VR (Virtual Reality) device, AR (Augmented Reality) device, etc. Server 103 can be a standalone physical server, a server cluster, a distributed system, or a cloud server providing cloud services. Network 110 can be a wired or wireless network, examples of which include, but are not limited to, the Internet, corporate intranet, local area network, wide area network, mobile communication network, and combinations thereof.

[0062] Currently, the mainstream super-resolution reconstruction method based on image scanning microscopy employs a pixel redistribution algorithm. This method is implemented through the following steps: Pixel redistribution is the mainstream super-resolution image reconstruction method based on image scanning microscopy, and its steps include: Peak detection: performing spot peak detection on the detector array image corresponding to each scanning point; Sub-pixel localization: performing Gaussian fitting or centroid calculation on the detected spot to determine its center with sub-pixel accuracy; Applying a digital pinhole: using a Gaussian function as a virtual pinhole to weight the spot, suppressing background and defocus light; Pixel redistribution: repositioning the weighted sub-pixel points into a higher resolution grid; Summation: accumulating the points that fall into the same high-resolution grid to generate the final super-resolution image.

[0063] The core algorithm of pixel-based reconstruction methods is point-by-point serial processing, which requires sequentially locating and redistributing thousands of focal points. This results in low computational efficiency and cannot meet the real-time imaging requirements of dynamic processes in living cells. The accuracy of sub-pixel localization heavily depends on the signal-to-noise ratio (SNR) of each focal spot. Under low illumination, low marker density, or rapid imaging conditions, a decrease in SNR can lead to localization errors, introduce artifacts, and significantly reduce reconstruction quality. The algorithm involves multiple parameters that require manual optimization, such as peak detection threshold, fitting algorithm, and digital pinhole size. This process is cumbersome, has low repeatability, and requires a high level of user experience. While suppressing the background, the digital pinhole also discards photons that may contain useful high-frequency information around the spot, limiting the potential for further resolution improvement.

[0064] Therefore, there is an urgent need in this field for a new super-resolution image reconstruction method that can overcome the above-mentioned shortcomings, has high speed, high robustness and simple process.

[0065] To address the problems of slow processing speed, sensitivity to signal-to-noise ratio, complex processes, and insufficient information utilization in the aforementioned pixel redistribution techniques, this invention provides a frequency decomposition-based image scanning microscopy super-resolution reconstruction method. It proposes a novel image scanning microscopy super-resolution reconstruction paradigm. Unlike the serial approach based on "point spread function localization and redistribution," this invention achieves direct super-resolution image synthesis without sub-pixel localization, digital pinholes, or Fourier transforms through frequency separation and parallel processing. This technical solution fundamentally solves the three core pain points of related technologies: slow processing speed, sensitivity to signal-to-noise ratio, and complex processes, providing a faster, more stable, and simpler super-resolution imaging tool.

[0066] According to an embodiment of the present invention, an embodiment of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0067] This embodiment provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition, which can be used in the aforementioned terminal device. Figure 2 This is a flowchart of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps:

[0068] Step S201: Obtain the set of original image data blocks; wherein, the set of original image data blocks includes multiple original dot matrix images obtained by performing two-dimensional scanning of the target sample using an image scanning microscopy system.

[0069] Specifically, the target sample is scanned using an image scanning microscopy system. At each scanning position, an array detector acquires the corresponding set of original image patches. In other words, the target sample is scanned in two dimensions using the image scanning microscopy system, and at each scanning position... The array detector captures a corresponding two-dimensional spot image, called an "image patch," which is the original raster image. The collection of all image patches constitutes the original dataset. .

[0070] The target samples can be microscopic biological or industrial materials, such as medical slides, cell samples, semiconductor wafers, and fiber materials.

[0071] Step S202: Perform multi-level frequency decomposition on the original dot matrix image, filter out the highest frequency noise component, and obtain multiple two-dimensional frequency components.

[0072] Specifically, each image block in the original image block set is processed. Using frequency separation technology, the low-frequency background information and high-frequency noise components in each image block are directly separated and filtered out. This step does not require complex Fourier transform and is completed directly in the spatial domain or through a specific envelope function, thus preserving the effective mid-to-high frequency structural information in the image block.

[0073] Furthermore, for each image patch Perform multi-level frequency decomposition, dividing each image block Decomposed into frequencies from high to low. Two-dimensional frequency components In this decomposition process, the highest frequency component The highest frequency noise component (i.e., the most frequent noise component) mainly contains random noise, which is directly discarded in subsequent processing to achieve front-end denoising. The remaining components cover effective information from background to details.

[0074] Step S203: Multiple two-dimensional frequency components are fused to generate a binarized digital pinhole mask.

[0075] Step S204: Based on the binarized digital pinhole mask and the highest two-dimensional frequency component among multiple two-dimensional frequency components, high-frequency information is extracted and synthesized to obtain a super-resolution reconstructed image.

[0076] Specifically, the binarized digital pinhole mask The high-frequency components containing super-resolution details obtained in step S202 are applied. The above (generally the next level of two-dimensional frequency components after discarding the highest frequency noise component, such as...) The super-resolution image blocks are obtained by stacking the super-resolution image blocks corresponding to each original bitmap image in parallel to obtain the super-resolution reconstructed image.

[0077] This embodiment provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition. By performing multi-level frequency decomposition on the original dot matrix image, the original dot matrix image is decomposed into multiple two-dimensional frequency components. These multiple two-dimensional frequency components are then fused to obtain a binarized digital pinhole mask. Through the standardized process of frequency decomposition-fusion-masking, the requirements for the operator's professional background are greatly reduced, improving the repeatability and ease of use of the method and expanding the application scenarios. Furthermore, based on the binarized digital pinhole mask and the highest two-dimensional frequency component among the multiple two-dimensional frequency components, the complete high-frequency component after adaptive masking is directly extracted and retained, making fuller use of the photon information collected by the detector. This theoretically has the potential to explore higher limit resolutions. Through the technical path of frequency separation and parallel processing, direct super-resolution image synthesis without sub-pixel positioning, digital pinhole, or Fourier transform is achieved, improving the processing speed of super-resolution imaging and enabling the super-resolution reconstructed image to have a high signal-to-noise ratio and high resolution.

[0078] This embodiment provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition, which can be used in the aforementioned terminal device. Figure 3 This is a flowchart of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps:

[0079] Step S301: Obtain the original image data block set; wherein, the original image data block set includes multiple original raster images obtained by performing two-dimensional scanning of the target sample using an image scanning microscopy system. For details, please refer to [link to relevant documentation]. Figure 2 Step S201 of the illustrated embodiment will not be described again here.

[0080] Step S302: Perform multi-level frequency decomposition on the original dot matrix image, filter out the highest frequency noise component, and obtain multiple two-dimensional frequency components.

[0081] Specifically, step S302 includes:

[0082] Step S3021: Frequency sieving is performed on the original dot matrix image to obtain candidate mode functions.

[0083] Among them, two-dimensional empirical mode decomposition is used to... Decomposed into several modal components from high frequency to low frequency, two-dimensional empirical mode decomposition is a completely data-driven and adaptive decomposition process. It separates different frequency components through "sieving". The process of two-dimensional empirical mode decomposition reflects the process of extracting the local highest and second highest frequencies.

[0084] In some optional implementations, step S3021 above includes:

[0085] Step a1: Initialize the original bitmap image to obtain local extrema.

[0086] Specifically, the original raster image is initialized, and local extrema (maximum and minimum points) are obtained. The specific steps include: obtaining the coordinates and pixel value of each pixel in the original raster image; selecting a neighborhood window according to the image resolution, and extracting all valid pixel values ​​within the neighborhood window for each pixel in the image; if the current pixel value is greater than or equal to all other valid pixel values ​​within the neighborhood window, the current pixel value is marked as a local maximum point; conversely, if the current pixel value is less than or equal to all other valid pixel values ​​within the neighborhood window, the current pixel value is marked as a local minimum point; after traversing all pixels, a set of local extrema points is obtained.

[0087] Step a2: Perform surface interpolation on the local extreme points to obtain the upper and lower envelope surfaces, and calculate the mean envelope surface based on the upper and lower envelope surfaces.

[0088] Specifically, surface interpolation is performed on local maxima and local minima to obtain upper and lower envelope surfaces, and then the mean envelope surface is calculated. The upper and lower envelope surfaces have the same size as the original point image.

[0089] Step a3: Calculate the difference between the original dot matrix image and the mean envelope to obtain the candidate mode function.

[0090] Specifically, the difference between the original bitmap image and the mean envelope is calculated to obtain the mode function. That is, the same pixel coordinates of the original bitmap image and the mean envelope are subtracted to obtain the mode function.

[0091] Step a4: Compare the candidate mode function with the mode function condition. If the candidate mode function satisfies the mode function condition, then output the candidate mode function.

[0092] Specifically, the screening range of the modal function (two-dimensional frequency components) is adjusted by the mean square error of the two envelope fittings. If the mean square error of the two envelope fittings is less than the set threshold, the screening stops.

[0093] Step S3022: Subtract the candidate mode function from the original dot matrix image to obtain the residual component.

[0094] Specifically, repeat steps a1-a3 above until the given modality function conditions are met to obtain the first modality function (i.e., candidate modality function). Subtract the first modality function from the original image to obtain the first residual component.

[0095] Step S3023: Iteratively decompose the residual components to obtain multiple mode functions.

[0096] Specifically, steps a1-a4 are repeated on the residual components to obtain the original bitmap image. One modal function.

[0097] Furthermore, the termination condition for iterative decomposition of the residual components is as follows: Determine the number of extreme points: Does the residual component contain at least 3 extreme points? If the condition is not met, stop the decomposition process; otherwise, continue repeating steps a1-a4 above. The number of modal functions obtained is the decomposition level. .

[0098] Furthermore, the fewer the number of extreme points of the residual components, the lower their frequency (approaching the stationary background components). That is, when there are fewer than 3 extreme points, the envelope cannot be effectively fitted, so the decomposition stops.

[0099] Furthermore, based on the statistical characteristics of local extrema, the number of layers (i.e., the number of decomposition levels) should be sufficient to cover the scale distribution range from high frequency to low frequency. Generally, when the last mode function after decomposition no longer contains obvious oscillations, it can be considered to have reached a suitable level. In most cases, an initial value (e.g., 4-6 layers) can be set experimentally or empirically. The screening range of the two-dimensional frequency components in each layer is adjusted by the mean square error of the two envelope fittings. A smaller mean square error value will lead to more screening times, which may extract purer frequency components, but will lead to an increase in the number of decompositions and smaller differences in the two-dimensional frequency components in each layer. A larger mean square error value will lead to fewer screening times, which may make the frequency components less pure and leave some modal mixing. In most cases, the mean square error value (e.g., 0.01-0.1) can be set experimentally or empirically.

[0100] Specifically, when the last residual term after decomposition no longer contains obvious oscillations, it is considered that... To reach a suitable level, it can also be set in advance. The initial value is set through experimentation or experience (e.g., 4-6).

[0101] Step S3024: Sort the multiple modal functions from high to low frequency to obtain multiple modal components.

[0102] Specifically, The modal functions are sorted from high to low frequency, denoted as . and a residual component The result can be written as:

[0103]

[0104] in, The number of frequencies is the number of decomposition levels.

[0105] Step S3025: Extract the highest frequency noise component from multiple modal components and filter out the highest frequency noise component to obtain multiple two-dimensional frequency components.

[0106] In some optional implementations, step S3025 above includes:

[0107] Step b1: Calculate the structural similarity index among multiple modal components, and determine the highest frequency noise component based on the structural similarity index.

[0108] Specifically, the determination of the highest frequency noise component includes: The structural similarity index is an indicator that measures the degree of similarity between two images; the structural similarity index is used to measure the original raster image. and each modal component after decomposition The similarity between them is determined by performing frequency decomposition on the original bitmap image and comparing the various modal components. The structural similarity index between them is used to determine whether each modal component is a noise-dominated component or a signal-dominated component; if the modal components If the structural similarity index is less than 0.90, the modal component is considered to be a noise-dominated component and should be discarded in subsequent execution.

[0109] Step b2 involves separating and filtering out the highest frequency noise component among the multiple modal components to obtain multiple two-dimensional frequency components.

[0110] Step S303 involves fusing multiple two-dimensional frequency components to generate a binarized digital pinhole mask. For details, please refer to [link to relevant documentation]. Figure 2 Step S203 of the illustrated embodiment will not be described again here.

[0111] Step S304: Based on the binarized digital pinhole mask and the highest two-dimensional frequency component among multiple two-dimensional frequency components, high-frequency information is extracted and synthesized to obtain the super-resolution reconstructed image. For details, please refer to [link to relevant documentation]. Figure 2 Step S204 of the illustrated embodiment will not be described again here.

[0112] This embodiment provides an image scanning microscopic super-resolution reconstruction method based on frequency decomposition. By iteratively filtering and separating the modal components from high frequency to low frequency in the original bitmap image, the method finally obtains modal functions and residual components of different frequencies. Following the separation logic of high frequency first and then low frequency, the modal components obtained first correspond to the highest frequency information of the image, and the subsequent components correspond to the mid-frequency texture and low-frequency background in turn. Finally, the residual components are the stable trend base of the image with clear layer boundaries. This method not only ensures the orderliness of frequency decomposition but also maximizes the preservation of local detail features of the original bitmap image, providing high-quality and clearly layered basic data for subsequent component processing and super-resolution reconstruction.

[0113] This embodiment provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition, which can be used in the aforementioned terminal device. Figure 4 This is a flowchart of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition according to an embodiment of the present invention, such as... Figure 4 As shown, the process includes the following steps:

[0114] Step S401: Obtain the original image data block set; wherein, the original image data block set includes multiple original raster images obtained by performing two-dimensional scanning of the target sample using an image scanning microscopy system. For details, please refer to [link to relevant documentation]. Figure 3 Step S301 of the illustrated embodiment will not be described again here.

[0115] Step S402 involves performing multi-level frequency decomposition on the original bitmap image, filtering out the highest frequency noise component, and obtaining multiple two-dimensional frequency components. For details, please refer to [link to relevant documentation]. Figure 3 Step S302 of the illustrated embodiment will not be described again here.

[0116] Step S403: Multiple two-dimensional frequency components are fused to generate a binarized digital pinhole mask.

[0117] Specifically, step S403 includes:

[0118] Step S4031: Component fusion of multiple two-dimensional frequency components is performed to obtain an intermediate image.

[0119] Specifically, from the two-dimensional frequency components obtained by decomposition, one or more mid-to-high frequency components (generally 1000-10 ... arrive , The two-dimensional frequency components are superimposed and fused to generate an intermediate image with a significantly improved signal-to-noise ratio. The intermediate image retains the morphological features of the light spot, but the background noise is effectively suppressed.

[0120] Step S4032: Based on the intermediate image, a solid circular structure is identified and extracted using an automatic detection algorithm for solid circular spots; wherein, the solid circular structure conforms to the diffraction-limited light spot characteristics.

[0121] Specifically, for the fused intermediate image An automatic detection algorithm for solid circular spots is applied. This algorithm, based on morphological operations (such as opening and closing operations) and connected component analysis, intelligently identifies spots of a specified diameter in the intermediate image that meet the characteristics of diffraction-limited spots. Solid circular structure within a pixel. The full width at half maximum (FWHM) of a Gaussian spot is theoretically limited by optical diffraction in the imaging system.

[0122] Step S4033: The pixel grayscale information corresponding to the solid circular structure is binarized to obtain a binarized digital pinhole mask.

[0123] Specifically, a global image thresholding method is used to binarize the pixel grayscale information within the detected solid circular structural region. Pixel grayscale values ​​greater than the threshold are set to 1, and otherwise to 0. The threshold is a global threshold and can be set to 0. This generates a corresponding binarized digital pinhole mask for each image block, and the binarized digital pinhole mask accurately marks the effective spatial distribution of the light spot.

[0124] Step S404: Based on the binarized digital pinhole mask and the highest two-dimensional frequency component among multiple two-dimensional frequency components, high-frequency information is extracted and synthesized to obtain the super-resolution reconstructed image. For details, please refer to [link to relevant documentation]. Figure 3 Step S304 of the illustrated embodiment will not be described again here.

[0125] This embodiment provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition. It fuses multiple two-dimensional frequency components to obtain an intermediate image, preserving the morphological features of the light spot, improving the signal-to-noise ratio, and effectively suppressing background noise. Furthermore, it utilizes an automatic detection algorithm for solid circular spots to identify and extract solid circular structures, accurately identifying diffraction-limited light spots, reducing manual intervention, and improving processing speed. Finally, it binarizes the pixel grayscale information corresponding to the solid circular structures, converting them into a binary digital pinhole mask. For light spots under different samples and imaging conditions, the mask can automatically adjust the number and distribution of pinholes without manual parameter modification. The binary digital pinhole mask dynamically adapts to the light spot features, supporting super-resolution reconstruction and solving the problems of poor image quality, difficult feature localization, and poor adaptability to subsequent processing.

[0126] This embodiment provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition, which can be used in the aforementioned terminal device. Figure 5 This is a flowchart of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition according to an embodiment of the present invention, such as... Figure 5 As shown, the process includes the following steps:

[0127] Step S501: Obtain the original image data block set; wherein, the original image data block set includes multiple original raster images obtained by performing two-dimensional scanning of the target sample using an image scanning microscopy system. For details, please refer to [link to relevant documentation]. Figure 4 Step S401 of the illustrated embodiment will not be described again here.

[0128] Step S502 involves performing multi-level frequency decomposition on the original bitmap image, filtering out the highest frequency noise component, and obtaining multiple two-dimensional frequency components. For details, please refer to [link to relevant documentation]. Figure 4 Step S402 of the illustrated embodiment will not be described again here.

[0129] Step S503 involves fusing multiple two-dimensional frequency components to generate a binarized digital pinhole mask. For details, please refer to [link to relevant documentation]. Figure 4 Step S403 of the illustrated embodiment will not be described again here.

[0130] Step S504: Based on the binarized digital pinhole mask and the highest two-dimensional frequency component among multiple two-dimensional frequency components, high-frequency information is extracted and synthesized to obtain a super-resolution reconstructed image.

[0131] Specifically, step S504 includes:

[0132] Step S5041: Multiply the binarized digital pinhole mask and the highest two-dimensional frequency component pixel by pixel to obtain a super-resolution image block.

[0133] Specifically, the binarized digital pinhole mask and the highest two-dimensional frequency component (i.e., the next level component after discarding the highest frequency noise component) are multiplied pixel by pixel. The high-frequency details corresponding to the real signal are "cut out" using the binarized digital pinhole mask, while the background area is set to zero. Finally, a single super-resolution image patch with background noise removed and resolution enhanced is obtained. , where pixel-by-pixel multiplication can be expressed as:

[0134]

[0135] Step S5042: Map the pixels corresponding to the super-resolution image blocks to the global grid in parallel to generate the initial super-resolution image.

[0136] Specifically, based on super-resolution image patches Corresponding original scan coordinates The image maps all its pixels in parallel to a global grid with a higher resolution (i.e. smaller pixel size). When pixels from multiple image patches are mapped to the same location on the global grid, their intensity values ​​are accumulated. Finally, by summing the values ​​across the entire grid, a complete initial super-resolution image is generated.

[0137] Step S5043: Perform image post-processing on the initial super-resolution image to obtain the super-resolution reconstructed image.

[0138] Specifically, the initial super-resolution image is subjected to necessary contrast stretching or normalization to optimize the visual effect and output the final super-resolution reconstructed image.

[0139] This embodiment provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition. By directly extracting and retaining the complete high-frequency components after adaptive masking, it makes fuller use of the photon information collected by the detector, thus theoretically possessing the potential to tap into higher limit resolution.

[0140] The following specific embodiment illustrates the steps of an image scanning microscopic super-resolution reconstruction method based on frequency decomposition.

[0141] Example 1:

[0142] like Figure 6 As shown, the super-resolution imaging algorithm based on Fourier transform and pixel redistribution includes the following steps: performing Fourier transform on m original image data sequentially to obtain a frequency domain image; performing frequency domain filtering on the frequency domain image; and then converting the processed frequency domain data back to the spatial domain through inverse Fourier transform; performing image opening operation on the inversely transformed image to further remove noise and refine the outline of bright spots, resulting in a clearer sample point image; utilizing the difference in the sampling point position of each image in the original data, mapping the bright spots of the m images onto a unified high-resolution pixel grid; and superimposing the repositioned m images to obtain a super-resolution image. It can be seen that the above-mentioned super-resolution imaging algorithm based on Fourier transform has obvious drawbacks such as frequency domain limitations, strong data dependence, and limited applicable scenarios. The algorithm requires sequentially performing Fourier transform, filtering, and inverse transform on each frame of the original low-resolution image, resulting in high computational complexity.

[0143] Furthermore, for m original data images, the wide-field resolution image generated by simply superimposing the original data using the relevant pixel redistribution method has a low resolution.

[0144] Therefore, this embodiment provides an image scanning microscopy super-resolution reconstruction method based on frequency decomposition, such as... Figure 7 As shown, the overall process of the image scanning microscopy super-resolution reconstruction method based on frequency decomposition includes:

[0145] like Figure 8As shown, in image reconstruction using image scanning microscopy, the original dot matrix image (i.e., m original data images) is decomposed into several two-dimensional frequency components from high frequency to low frequency. The highest frequency component corresponds to high frequency noise in the image and is discarded in subsequent calculations.

[0146] After fusing several mid- and high-frequency components from several two-dimensional frequency components, an intermediate image with a significantly improved signal-to-noise ratio is generated.

[0147] Subsequently, using an automatic detection algorithm for solid circular spots, solid circular structures that conform to the diffraction-limited spot characteristics are intelligently identified and extracted from the fused image, and then converted into an adaptive binary digital pinhole mask.

[0148] By applying a binarized digital pinhole mask to the corresponding sub-high frequency components, high-frequency information representing super-resolution details can be accurately preserved and enhanced while completely suppressing background noise.

[0149] Finally, all the processed raster images are stacked in parallel to obtain a high signal-to-noise ratio and high resolution reconstruction result, namely a super-resolution reconstructed image.

[0150] The above-described embodiment 1 has the following advantages:

[0151] 1) It completely avoids the complex steps required in pixel redistribution methods that are highly dependent on human experience, such as peak detection, subpixel fitting (e.g. Gaussian fitting), and fixed parameter digital pinhole settings. The entire process is highly automated. Through the standardized process of "frequency decomposition-fusion-mask", it greatly reduces the requirements for the operator's professional background, improves the repeatability and ease of use of the method, and is conducive to the promotion and standardized application of the technology.

[0152] 2) In the relevant pixel redistribution method, while suppressing the background, the digital pinhole will inevitably discard the useful high-frequency information around the spot. Therefore, this embodiment makes fuller use of the photon information collected by the detector by directly extracting and retaining the complete high-frequency components after adaptive masking, thus theoretically having the potential to tap higher limit resolution.

[0153] In summary, this embodiment, through a fundamental shift in technological paradigm, successfully addresses the three core pain points of pixel redistribution methods that have long existed: slow processing speed, sensitivity to signal-to-noise ratio, and complex processes. It provides a super-resolution imaging tool that achieves a leap forward in speed, robustness, and ease of use, powerfully promoting the in-depth application of image scanning microscopy technology in fields such as life science research and industrial inspection.

[0154] This embodiment also provides an image scanning microscopy super-resolution reconstruction apparatus based on frequency decomposition, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0155] This embodiment provides an image scanning microscopic super-resolution reconstruction device based on frequency decomposition, such as... Figure 9 As shown, it includes:

[0156] The acquisition module 901 is used to acquire a set of raw image data blocks; wherein, the set of raw image data blocks includes multiple raw dot matrix images obtained by performing two-dimensional scanning of the target sample using an image scanning microscopy system;

[0157] The multi-level frequency decomposition module 902 is used to perform multi-level frequency decomposition on the original dot matrix image, filter out the highest frequency noise component, and obtain multiple two-dimensional frequency components.

[0158] The fusion processing module 903 is used to fuse multiple two-dimensional frequency components to generate a binary digital pinhole mask.

[0159] The extraction and synthesis module 904 is used to extract and synthesize high-frequency information based on the binarized digital pinhole mask and the highest two-dimensional frequency component among multiple two-dimensional frequency components to obtain a super-resolution reconstructed image.

[0160] In some optional implementations, the multi-level frequency decomposition module 902 includes:

[0161] A screening unit is used to perform frequency screening on the original dot matrix image to obtain candidate mode functions;

[0162] The calculation unit is used to subtract the candidate mode function from the original dot matrix image to obtain the residual component;

[0163] An iterative decomposition unit is used to iteratively decompose the residual components to obtain multiple mode functions;

[0164] The sorting unit is used to sort the multiple modal functions from high to low frequency to obtain multiple modal components;

[0165] The filtering unit is used to extract the highest frequency noise component from the multiple modal components and filter out the highest frequency noise component to obtain multiple two-dimensional frequency components.

[0166] In some alternative implementations, the screening unit includes:

[0167] An initialization subunit is used to initialize the original dot matrix image and obtain local extreme points;

[0168] An interpolation subunit is used to perform surface interpolation on the local extreme points to obtain an upper envelope surface and a lower envelope surface, and to calculate a mean envelope surface based on the upper envelope surface and the lower envelope surface.

[0169] A computational subunit is used to calculate the difference between the original dot matrix image and the mean envelope surface to obtain the candidate mode function;

[0170] The comparison subunit is used to compare the candidate mode function with the mode function condition. If the candidate mode function satisfies the mode function condition, the candidate mode function is output.

[0171] In some optional implementations, the filtering unit includes:

[0172] A subunit is determined for calculating the structural similarity index between multiple modal components, and the highest frequency noise component is determined based on the structural similarity index.

[0173] The separation and filtering subunit is used to separate and filter out the highest frequency noise component among the multiple modal components to obtain multiple two-dimensional frequency components.

[0174] In some alternative implementations, the fusion processing module 903 includes:

[0175] A fusion unit is used to fuse multiple two-dimensional frequency components to obtain an intermediate image.

[0176] The extraction unit is used to identify and extract solid circular structures based on the intermediate image using an automatic detection algorithm for solid circular spots; wherein the solid circular structures conform to the diffraction-limited spot characteristics;

[0177] The binarization processing unit is used to binarize the pixel grayscale information corresponding to the solid circular structure to obtain the binarized digital pinhole mask.

[0178] In some alternative implementations, the extraction and synthesis module 904 includes:

[0179] The multiplication unit is used to multiply the binarized digital pinhole mask and the highest two-dimensional frequency component pixel by pixel to obtain a super-resolution image block.

[0180] A mapping unit is used to map the pixels corresponding to the super-resolution image block to a global grid in parallel to generate an initial super-resolution image;

[0181] An image post-processing unit is used to perform image post-processing on the initial super-resolution image to obtain the super-resolution reconstructed image.

[0182] The image scanning microscopy super-resolution reconstruction apparatus based on frequency decomposition provided in this embodiment of the invention can execute the image scanning microscopy super-resolution reconstruction method based on frequency decomposition provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0183] Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0184] The following is a detailed reference. Figure 10 This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 1001, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from memory 1008 into random access memory (RAM) 1003. The RAM 1003 also stores various programs and data required for the operation of the electronic device. The processor 1001, ROM 1002, and RAM 1003 are interconnected via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0185] Typically, the following devices can be connected to the I / O interface 1005: input devices 1006 including, for example, a touchscreen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 1007 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; memory devices 1008 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows electronic devices to exchange data via wireless or wired communication with other devices. Although Figure 10 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0186] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 1009, or installed from a memory 1008, or installed from a ROM 1002. When the computer program is executed by the processor 1001, it performs the functions defined in the frequency decomposition-based image scanning microscopy super-resolution reconstruction method of the present invention.

[0187] Figure 10 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0188] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, it implements the image scanning microscopy super-resolution reconstruction method based on frequency decomposition shown in the above embodiments.

[0189] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0190] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for super-resolution reconstruction of images using scanning microscopy based on frequency decomposition, characterized in that, The method includes: Obtain a set of raw image data blocks; wherein, the set of raw image data blocks includes multiple raw dot matrix images obtained by performing two-dimensional scanning of the target sample using an image scanning microscopy system; The original dot matrix image is subjected to multi-level frequency decomposition to filter out the highest frequency noise component, resulting in multiple two-dimensional frequency components. The multiple two-dimensional frequency components are fused to generate a binary digital pinhole mask. Based on the binarized digital pinhole mask and the highest two-dimensional frequency component among the multiple two-dimensional frequency components, high-frequency information is extracted and synthesized to obtain a super-resolution reconstructed image. The process involves performing multi-level frequency decomposition on the original dot matrix image, filtering out the highest frequency noise component, and obtaining multiple two-dimensional frequency components, including: The original dot matrix image is subjected to frequency sieving to obtain candidate mode functions; Subtract the candidate mode function from the original dot matrix image to obtain the residual components; The residual components are iteratively decomposed to obtain multiple mode functions; The multiple modal functions are sorted from high to low frequency to obtain multiple modal components; The highest frequency noise component is extracted from the multiple modal components, and the highest frequency noise component is filtered out to obtain multiple two-dimensional frequency components; The step of extracting the highest frequency noise component from the multiple modal components and filtering out the highest frequency noise component to obtain multiple two-dimensional frequency components includes: Calculate the structural similarity index among the multiple modal components, and determine the highest frequency noise component based on the structural similarity index; The highest frequency noise component among the multiple modal components is separated and filtered out to obtain multiple two-dimensional frequency components.

2. The method according to claim 1, characterized in that, The step of frequency screening of the original dot matrix image to obtain candidate mode functions includes: The original dot matrix image is initialized to obtain local extreme points; Surface interpolation is performed on the local extreme points to obtain the upper envelope and the lower envelope, and the mean envelope is calculated based on the upper envelope and the lower envelope. The difference between the original dot matrix image and the mean envelope is calculated to obtain the candidate mode function; The candidate mode function is compared with the mode function condition. If the candidate mode function satisfies the mode function condition, the candidate mode function is output.

3. The method according to claim 1, characterized in that, The step of fusing multiple two-dimensional frequency components to generate a binary digital pinhole mask includes: The multiple two-dimensional frequency components are fused together to obtain an intermediate image; Based on the intermediate image, a solid circular structure is identified and extracted using an automatic detection algorithm for solid circular spots; wherein the solid circular structure conforms to the characteristics of a diffraction-limited light spot. The pixel grayscale information corresponding to the solid circular structure is binarized to obtain the binarized digital pinhole mask.

4. The method according to claim 1, characterized in that, The process of extracting and synthesizing high-frequency information based on the binarized digital pinhole mask and the highest two-dimensional frequency component among multiple two-dimensional frequency components to obtain a super-resolution reconstructed image includes: The binarized digital pinhole mask and the highest two-dimensional frequency component are multiplied pixel by pixel to obtain a super-resolution image block; The pixels corresponding to the super-resolution image blocks are mapped in parallel to the global grid to generate an initial super-resolution image; The initial super-resolution image is post-processed to obtain the super-resolution reconstructed image.

5. An image scanning microscopic super-resolution reconstruction device based on frequency decomposition, characterized in that, The device includes: An acquisition module is used to acquire a set of raw image data blocks; wherein, the set of raw image data blocks includes multiple raw dot matrix images obtained by performing two-dimensional scanning of the target sample using an image scanning microscopy system; The multi-level frequency decomposition module is used to perform multi-level frequency decomposition on the original dot matrix image, filter out the highest frequency noise component, and obtain multiple two-dimensional frequency components. The fusion processing module is used to fuse multiple two-dimensional frequency components to generate a binary digital pinhole mask. The extraction and synthesis module is used to extract and synthesize high-frequency information based on the binarized digital pinhole mask and the highest two-dimensional frequency component among the multiple two-dimensional frequency components to obtain a super-resolution reconstructed image. The multi-level frequency decomposition module includes: A screening unit is used to perform frequency screening on the original dot matrix image to obtain candidate mode functions; The calculation unit is used to subtract the candidate mode function from the original dot matrix image to obtain the residual component; An iterative decomposition unit is used to iteratively decompose the residual components to obtain multiple mode functions; The sorting unit is used to sort the multiple modal functions from high to low frequency to obtain multiple modal components; A filtering unit is used to extract the highest frequency noise component from the multiple modal components and filter out the highest frequency noise component to obtain multiple two-dimensional frequency components. The filtration unit includes: A subunit is determined for calculating the structural similarity index between multiple modal components, and the highest frequency noise component is determined based on the structural similarity index. The separation and filtering subunit is used to separate and filter out the highest frequency noise component among the multiple modal components to obtain multiple two-dimensional frequency components.

6. An electronic device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the image scanning microscopic super-resolution reconstruction method based on frequency decomposition as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the image scanning microscopic super-resolution reconstruction method based on frequency decomposition as described in any one of claims 1 to 4.

8. A computer program product, characterized in that, Includes computer instructions for causing a computer to perform the image scanning microscopic super-resolution reconstruction method based on frequency decomposition as described in any one of claims 1 to 4.

Images