Digital image processing methods, devices and related components for improving STED imaging quality

Abstract

This invention provides a digital image processing method, apparatus, and related components for improving the quality of STED imaging. The method involves imaging fluorescent beads using an STED microscope, performing a first image processing step, Fourier transform processing, and cropping processing on the image to obtain an FFT image. Then, the noisy FFT image undergoes thresholding, inverse Fourier transform, and a second image processing step to obtain the final STED image, which has reduced background noise and improved image resolution. This method is simple, requires no additional hardware modifications, and is also applicable to complex structures such as mitochondria.

Description

Technical Field

[0001] This invention relates to the field of optical microscope imaging technology, and in particular to digital image processing methods, apparatus and related components for improving the imaging quality of STED (Scanning Microscope) images. Background Technology

[0002] Stimulated emission depletion microscopy (STED) uses two laser beams to simultaneously illuminate the sample. One laser beam serves as the excitation beam, exciting fluorescent molecules within the focal region to an excited state. The other laser beam is a ring laser with zero intensity at its center, used to suppress fluorescence in fluorophores around the excitation focal point, thus achieving super-resolution. However, problems such as photobleaching, photodamage, and STED-specific background noise exist during STED microscopy. The main background noise sources include direct excitation noise (DEN) and incomplete depletion noise (IDN).

[0003] DEN is generated when fluorophores are unintentionally excited by excitation light, and IDN is caused when fluorescence leaks from the region around the ring laser that is not completely depleted. Both types of noise are affected by the intensity of the STED laser, so it is necessary to suppress both DEN and IDN. However, the suppression methods are quite complex, especially in 3D DTED imaging. Furthermore, the additional loss required to achieve 3D super-resolution will reduce the signal-to-background ratio (SBR). Therefore, it is necessary to reduce STED-specific background noise to improve the image SBR and resolution.

[0004] Existing technologies for reducing STED-specific background noise include: 1. STED subtraction, primarily targeting direct excitation noise (DEN), which uses only the STED laser beam to scan the sample to capture background information, then subtracts the background information from the STED image; 2. Digital enhancement, which improves the resolution of fluorescent beads and microtube samples by digitally enhancing the eraser power, but this results in information loss for complex samples (such as mitochondria). However, these methods do not address the issue of IDN. Existing technologies suppress DEN by adjusting the STED intensity during imaging using lifetime-tuned photon separation, or by estimating and subtracting background noise using the Gaussian distribution of stimulated emission double depletion technology and a second STED pulse detected by time correlation. However, this approach requires precise time alignment, time correlation measurements, and careful optimization of the second STED pulse intensity and subtraction factor, making it very complex. Furthermore, polarization switching can also accurately subtract STED-specific background noise (including DEN and IDN). In polarization-switched STED, reversing the polarization of the STED beam produces a center-filled donut pattern, effectively capturing and subtracting background information. Dynamic polarization switching can rapidly reduce background noise. However, existing methods for reducing background noise in STED microscopy either struggle to handle complex structures such as mitochondria or require additional hardware modifications, making them complex to operate. Summary of the Invention

[0005] The present invention provides a digital image processing method, apparatus and related components for improving the imaging quality of STED microscopes, aiming to solve the problem that existing methods for reducing background noise in STED microscopes are difficult to handle complex structures such as mitochondria or require additional hardware modifications, which leads to operational complexity.

[0006] In a first aspect, embodiments of the present invention provide a digital image processing method for improving STED imaging quality, comprising:

[0007] The fluorescent beads were imaged using a STED microscope to obtain initial STED images and co-aggregation images;

[0008] The initial STED image and co-aggregated image are subjected to the first image processing to obtain the actual loss light image. The actual loss light image is then subjected to Fourier transform processing and truncation processing to obtain the FFT image.

[0009] The FFT image is compared with the ideal loss light FFT image, and it is determined whether there is noise in the FFT image. If there is noise, the FFT image is thresholded, and then the thresholded FFT image is subjected to inverse Fourier transform to obtain the ideal loss light image.

[0010] The initial STED image and the ideal loss light image are subjected to a second image processing to obtain the final STED image.

[0011] Secondly, embodiments of the present invention provide a digital image processing apparatus for improving STED imaging quality, comprising:

[0012] The imaging unit is used to image fluorescent beads using a STED microscope to obtain initial STED images and co-aggregation images.

[0013] The first processing unit is used to perform a first image processing on the initial STED image and the co-aggregation image to obtain an actual loss light image, and to perform Fourier transform processing and truncation processing on the actual loss light image to obtain an FFT image.

[0014] The noise determination unit is used to compare the FFT image with the ideal loss light FFT image and determine whether there is noise in the FFT image. If there is noise, the FFT image is thresholded and then the thresholded FFT image is subjected to inverse Fourier transform to obtain the ideal loss light image.

[0015] The second processing unit is used to perform a second image processing on the initial STED image and the ideal loss light image to obtain the final STED image.

[0016] Thirdly, embodiments of the present invention 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 digital image processing method as described above.

[0017] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the digital image processing method as described above.

[0018] This invention provides a digital image processing method, apparatus, and related components for improving the quality of STED imaging. The method involves imaging fluorescent beads using an STED microscope, performing a first image processing step, Fourier transform processing, and cropping processing on the image to obtain an FFT image. Then, the noisy FFT image undergoes thresholding, inverse Fourier transform, and a second image processing step to obtain the final STED image, which has reduced background noise and improved image resolution. This method is simple, requires no additional hardware modifications, and is also applicable to complex structures such as mitochondria. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating a digital image processing method for improving STED imaging quality provided in an embodiment of the present invention;

[0021] Figure 2 A schematic diagram of a sub-process of a digital image processing method for improving STED imaging quality provided in an embodiment of the present invention;

[0022] Figure 3 This is a schematic diagram of another sub-process of a digital image processing method for improving STED imaging quality provided in an embodiment of the present invention;

[0023] Figure 4 This is a schematic diagram of another sub-process of a digital image processing method for improving STED imaging quality provided in an embodiment of the present invention;

[0024] Figure 5 This is a schematic diagram of another sub-process of a digital image processing method for improving STED imaging quality provided in an embodiment of the present invention;

[0025] Figure 6 This is a schematic diagram of another sub-process of a digital image processing method for improving STED imaging quality provided in an embodiment of the present invention;

[0026] Figure 7 A simulation diagram illustrating a digital image processing method for improving STED imaging quality provided in an embodiment of the present invention;

[0027] Figure 8 This is a schematic block diagram of a digital image processing device for improving STED imaging quality, provided as an embodiment of the present invention. Detailed Implementation

[0028] 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, not all, of the embodiments of the present invention. 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.

[0029] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0030] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0031] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0032] Please see Figure 1 This invention provides a digital image processing method for improving STED imaging quality, including steps S10-S40:

[0033] S10. Use an STED microscope to image the fluorescent beads to obtain the initial STED image and co-aggregation image;

[0034] In this step, fluorescent beads are typically tiny particles labeled with fluorescent dyes that fluoresce under a STED microscope. The STED microscope works by using two main laser beams: an excitation beam to excite fluorescent molecules and a suppression beam (STED beam, also known as a ring beam) to bring some fluorescent molecules back to their ground state through stimulated emission. By using the STED microscope, only fluorescent molecules located in the region at the center of the excitation spot can fluoresce, thus achieving high-resolution imaging.

[0035] When imaging fluorescent beads using a STED microscope, a high-resolution STED image (i.e., the initial STED image) can be obtained by adjusting the parameters of the excitation beam and the STED beam. In this image, the edges and internal structure of the fluorescent beads are very clear because the STED technique effectively suppresses the fluorescence signal in the surrounding area, highlighting the details in the central region. Further adjustments to the STED microscope settings or the use of different imaging modes, followed by the use of a laser beam to excite the fluorescent molecules, yield a co-focal image. This image shows the overall shape and position of the fluorescent beads, but the details and resolution are not as clear as the initial STED image. This embodiment uses an STED microscope to image fluorescent beads, obtaining both an initial STED image and a confocal image. These two images provide different information about the structure and properties of the fluorescent beads for subsequent implementation processes.

[0036] S20. Perform the first image processing on the initial STED image and the co-aggregation image to obtain the actual loss light image. Perform Fourier transform processing and truncation processing on the actual loss light image to obtain the FFT image.

[0037] In this step, during the imaging of fluorescent beads using a STED microscope, the STED beam reduces the effective fluorescence emission area, thereby improving resolution. However, excessively high power of the STED beam or other factors can cause light loss, leading to a decrease in image quality, such as reduced image resolution, blurring, or decreased contrast. Therefore, the initial STED image and co-aggregated image need to be processed first to obtain the actual loss light image. Then, the actual loss light image is processed by Fourier transform. The Fourier transform process uses the Fourier transform algorithm to convert the spatial domain (or time domain) signal of the actual loss light image into a frequency domain signal, thereby obtaining the frequency domain of the actual loss light image. In the frequency domain, the distribution of different frequencies in the actual loss light image can be observed, such as low-frequency and high-frequency distributions. Then, the frequencies of interest are truncated to remove unwanted frequency components, thus obtaining the FFT image.

[0038] In one embodiment, such as Figure 2 As shown, S20 includes:

[0039] S21. Compare the initial STED image and the co-aggregated image to obtain the comparison result;

[0040] S22. If the comparison result shows a difference, the co-aggregated image is subtracted from the initial STED image to obtain the actual loss light image.

[0041] In this embodiment, in S21, the initial STED image and the co-aggregated image need to be preprocessed. The preprocessing process includes denoising and image registration. For denoising, i.e. removing background noise from the image, it can be done by using a filter. For image registration, since the initial STED image and the co-aggregated image may come from different imaging systems or different imaging conditions, it is necessary to align them using image registration technology to ensure their consistency in spatial location. Then, the initial STED image and the co-aggregated image are compared to determine whether they are consistent in terms of resolution, size, and pixel depth. If they are inconsistent, it indicates that there is a difference. Alternatively, the gray values ​​or fluorescence intensities of the initial STED image and the co-aggregated image can be compared pixel by pixel, and then the difference between the two images can be calculated. The difference can be quantified by indicators such as mean square error (MSE) and structural similarity (SSIM). Based on the comparison results, the degree of difference between the two images is recorded. If the degree of difference exceeds a set threshold (e.g., MSE is greater than a certain value or SSIM is less than a certain value), it is considered that there is a difference.

[0042] In S22, image processing software or code is used to perform pixel-by-pixel subtraction between the initial STED image and the co-aggregated image that have differences.

[0043] In practice, the actual loss light image can be verified to ensure that it reflects the difference between the initial STED image and the co-aggregation image. Visualization tools (such as image viewers) can be used to check whether the actual loss light image contains the expected loss light information.

[0044] In one embodiment, such as Figure 3 As shown, S20 also includes:

[0045] S23. Perform Fourier transform processing on the actual loss light image to obtain the actual loss light FFT image;

[0046] S24. Use image software to extract the image region with a predetermined amplitude value from the actual loss light FFT image to obtain the FFT image.

[0047] In this embodiment, in S23, the actual lost light image is imported into image processing software that supports Fourier transform processing, such as MATLAB, Python (using NumPy and SciPy libraries), or professional image processing software (such as ImageJ, Photoshop plugins, etc.). Then, the function to perform Fourier transform is selected. For two-dimensional images, the two-dimensional FFT function is usually used. The software will calculate the FFT of the actual lost light image and generate the actual lost light FFT image. This image shows the representation of the actual lost light image in the frequency domain, where the high-frequency components are located near the center of the image, while the low-frequency components are located at the edges of the image.

[0048] In S24, the predetermined amplitude value can be determined based on the amplitude of noise in the actual lost light FFT image, the amplitude range of the signal of interest, etc. A selection tool (such as a rectangular selection box, a circular selection box, etc.) is used to extract the image region of the predetermined amplitude value in the actual lost light FFT image. This region can be a region with an amplitude value greater than, less than or equal to the predetermined amplitude value, thereby obtaining an FFT image. The FFT image includes amplitude information and phase information, which can be used for subsequent noise judgment.

[0049] In practical implementation, a mask technique can also be used to extract an image region with a predetermined amplitude value from the actual lossy light FFT image. The image region with the predetermined amplitude value is taken as the region of interest, and a two-dimensional array of the same size as the actual lossy light FFT image is created as a mask. In the mask, the element values ​​of the region of interest are set to 1, and the element values ​​of other regions are set to 0. The mask is multiplied element-wise with the actual lossy light FFT image, and the resulting image is the FFT image.

[0050] S30. Compare the FFT image with the ideal loss light FFT image and determine whether there is noise in the FFT image. If there is noise, perform thresholding on the FFT image and then perform inverse Fourier transform on the thresholded FFT image to obtain the ideal loss light image.

[0051] In this step, the ideal loss light FFT image is a predefined, noise-free FFT image representing the ideal loss light distribution. The FFT image is compared with the ideal loss light FFT image to determine if noise exists. Noise may manifest as high-frequency or low-frequency components, randomly distributed bright or dark spots, or regions significantly different from the ideal image. If noise is present, denoising is required. This step uses thresholding. Based on the noise detection results, a suitable threshold is set. This threshold can be global (applicable to the entire FFT image) or local (different thresholds are set for different regions of the image). Then, pixels with amplitude values ​​below or above the threshold are set to zero or subjected to other processing (such as removal, retention, or reduction of amplitude) until the FFT image is free of noise. Finally, the threshold-processed FFT image undergoes an inverse Fourier transform to obtain the ideal loss light image.

[0052] In one embodiment, such as Figure 4 As shown, S30 includes:

[0053] S31. Obtain the amplitude value of the FFT image, and calculate the spectrum of the FFT image based on the amplitude value;

[0054] S32. Compare the spectrum of the FFT image with the spectrum of the ideal loss light FFT image, and determine whether a specific region in the spectrum of the FFT image is 0. If it is not 0, then it is confirmed that the FFT image contains noise.

[0055] This embodiment describes the first method of noise assessment: acquiring the amplitude values ​​of the FFT image. These amplitude values ​​represent the intensity of the FFT image at different frequency components. Based on the amplitude values, the spectrum of the FFT image is calculated, typically displayed as a graph showing the relationship between frequency and corresponding amplitude values. Then, the spectrum of the ideal lossy light FFT image is acquired. This spectrum represents the frequency domain characteristics of the lossy light generated by the system in the absence of a fluorescence signal. The spectrum of the ideal lossy light FFT image is 0 in a specific region, known as the "donut" region in the center. The spectrum of the FFT image is compared with that of the ideal lossy light FFT image to determine if the spectrum of the specific region in the FFT image corresponds to the same region in the ideal lossy light FFT image. If it is 0, the FFT image is confirmed to be noise-free; otherwise, noise is confirmed. This process continues until the FFT image is free of noise before proceeding to the next step.

[0056] In one embodiment, such as Figure 5 As shown, S30 also includes:

[0057] S33. Obtain the phase information of the FFT image, and calculate the power spectrum of the FFT image based on the phase information;

[0058] S34. Determine whether one or more values ​​in the power spectrum are 0. If they are not 0, then confirm that the FFT image contains noise.

[0059] This embodiment describes the second method for noise assessment, which involves acquiring the phase information of the FFT image. This phase information represents the phase shift of each frequency component relative to a certain reference point. Based on the phase information, the power spectrum of the FFT image is calculated.

[0060] The power spectrum is calculated using the amplitude and phase information of the FFT image. Then, the presence of noise in the FFT image is determined. Before this determination, a reasonable power threshold needs to be set. This is because, in actual calculations, due to limitations in numerical precision, the values ​​in the power spectrum are unlikely to be completely zero. Therefore, the power threshold should be set to a sufficiently small positive number to distinguish between non-zero power caused by noise and the minute values ​​resulting from computational inaccuracies. The entire power spectrum is traversed, and the power value corresponding to each frequency component is checked. Noise is then assessed. If one or more frequency components are found to have power values ​​greater than the threshold (i.e., not zero in the sense of the threshold), noise is considered to exist in the FFT image. This is because noise typically produces non-zero power across multiple frequency components. Further operations are only performed after the FFT image is found to be free of noise.

[0061] In one embodiment, such as Figure 6 As shown, S30 also includes:

[0062] S35. Obtain the previous FFT image and filter the FFT image to obtain the filtered FFT image;

[0063] S36. Determine whether there is noise in the filtered FFT image. If there is noise, continue filtering the current FFT image. If there is no noise, perform an inverse Fourier transform on the current FFT image to obtain an ideal loss light image.

[0064] In this embodiment, the previous FFT image is the previous version of the current FFT image obtained from previous processing steps. Specifically, if the current FFT image is not filtered, the previous FFT image is the original FFT image (i.e., the FFT image of the first round); if the current FFT image is filtered, the previous FFT image is the FFT image after the previous round of filtering. The filtering methods include low-pass filtering, high-pass filtering, band-pass filtering, and band-stop filtering. To reduce high-frequency noise, low-pass filtering can be selected, that is, filtering the portion of the acquired FFT image below a predetermined frequency value to obtain a filtered FFT image. It is then determined whether the filtered FFT image contains noise. If the filtered FFT image still contains noise, and the noise level is higher than an acceptable threshold, the filtering effect is considered unsatisfactory. The filtering threshold is then adjusted, and filtering continues until the noise in the filtered FFT image is reduced to an acceptable level. Next, an inverse Fourier transform is performed on the current FFT image, and the IFFT algorithm is used to convert the filtered FFT image back to a time-domain image, that is, to perform an inverse transform on the complex amplitude values ​​of the FFT image to obtain an ideal lossy light image.

[0065] S40. Perform a second image processing on the initial STED image and the ideal loss light image to obtain the final STED image.

[0066] In this step, the ideal lossy light image represents the lossy light distribution generated by the STED microscope system in the absence of target fluorescence signal. The initial STED image and the ideal lossy light image are then preprocessed. This preprocessing includes image registration, where an image registration algorithm automatically aligns the initial STED image and the ideal lossy light image to ensure they are spatially perfectly matched. This includes operations such as translation, rotation, and scaling to eliminate any possible displacement or distortion. A second image processing step is then performed on the two images. This second image processing can be image subtraction. The specific process of image subtraction involves subtracting the ideal lossy light image from the initial STED image to obtain the final STED image. In image processing software or a programming environment, each pixel value of the initial STED image is subtracted from the corresponding pixel value of the ideal lossy light image; the resulting difference image is the final STED image. The final STED image represents the target fluorescence signal distribution after removing the lossy light signal.

[0067] The various images obtained from the entire processing method can be referenced. Figure 7 As shown, Figure 7 In the diagram, Figure A shows the STED donut, which represents a specific region where the STED value is not zero; Figure A also shows the FFT, which represents a specific region within the red dashed circle outline, which is a non-zero region; Figure B shows the Ideal donut, which represents an ideal loss light image with a specific region of zero (similar to a "donut" pattern); Figure B also shows the FFT, which represents a specific region within the red dashed circle outline, which is a zero region; Figure C shows the Confocal image after imaging the fluorescent beads; Figure D shows the initial STED image after imaging the fluorescent beads; Figure E shows the Donutbeam, which is the actual loss light image; Figure E also shows the FFT, which is the FFT image captured from the actual loss light (Nonzero indicates that a specific region of the captured FFT image is not zero); Figure F shows the Ideal donut beam, which represents the ideal loss light image; Figure F shows the Filtered FFT, which represents the filtered FFT image (Almost zero indicates that a specific region of the filtered FFT image is almost zero); Figure G shows the initial STED image after imaging the fluorescent beads; and Figure G shows the Redepleted STED, which is the final STED image.

[0068] In practical implementation, this processing method can be used to improve the resolution and signal-to-background ratio of dual organelle structures. Specifically, the STED images of the Golgi apparatus and nuclear pore complex are processed by deconvolution of the STED images using Huygens software. A ring light image is obtained from the confocal image and the STED image. Then, the ring light image is digitally enhanced by multiplying it by a digital power enhancement factor (K). At this time, the power of the ring light is digitally increased by a factor of K. The final STED image is obtained by subtracting the digitally enhanced ring light from the confocal image. Then, the image resolution of the final STED image is calculated using Fourier rings. This resolution is improved from the original 70nm to 35nm. This method also improves the signal-to-background ratio of the STED image of dual organelle structures.

[0069] Furthermore, this processing method can also improve the resolution and signal-to-background ratio of 3D-STED images. Specifically, by performing the aforementioned series of processing steps on the 3D-STED image of the fluorescent beads, and then performing image verification, it was found that the final STED image significantly eliminated background noise. By characterizing the resolution in the z-axis using the full width at half maximum (FWHM), the z-axis resolution was improved from 150 nm to 100 nm.

[0070] This invention also provides a digital image processing apparatus for improving STED imaging quality, which is used to execute any embodiment of the aforementioned digital image processing method for improving STED imaging quality. Specifically, please refer to... Figure 8 , Figure 8 This is a schematic block diagram of a digital image processing apparatus 500 for improving STED imaging quality according to an embodiment of the present invention. The digital image processing apparatus 500 for improving STED imaging quality includes:

[0071] Imaging unit 510 is used to image fluorescent beads using an STED microscope to obtain initial STED images and co-aggregation images;

[0072] The first processing unit 520 is used to perform a first image processing on the initial STED image and the co-aggregation image to obtain an actual loss light image, and to perform Fourier transform processing and truncation processing on the actual loss light image to obtain an FFT image.

[0073] The noise judgment unit 530 is used to compare the FFT image with the ideal loss light FFT image and determine whether there is noise in the FFT image. If there is noise, the FFT image is thresholded and then the thresholded FFT image is subjected to inverse Fourier transform to obtain the ideal loss light image.

[0074] The second processing unit 540 is used to perform a second image processing on the initial STED image and the ideal loss light image to obtain the final STED image.

[0075] In one embodiment, the first processing unit 520 is used to:

[0076] The initial STED image and the co-aggregated image are compared to obtain the comparison results;

[0077] If the comparison results show a difference, the co-aggregated image is subtracted from the initial STED image to obtain the actual loss light image.

[0078] In one embodiment, the first processing unit 520 is further configured to:

[0079] Perform Fourier transform processing on the actual loss light image to obtain the actual loss light FFT image;

[0080] An image region with a predetermined amplitude value is extracted from the actual loss light FFT image using image processing software to obtain the FFT image.

[0081] In one embodiment, the noise determination unit 530 is used for:

[0082] Obtain the amplitude value of the FFT image, and calculate the spectrum of the FFT image based on the amplitude value;

[0083] The spectrum of the FFT image is compared with the spectrum of the ideal loss light FFT image to determine whether a specific region in the spectrum of the FFT image is 0. If it is not 0, then the FFT image is confirmed to have noise.

[0084] In one embodiment, the noise determination unit 530 is further configured to:

[0085] The phase information of the FFT image is obtained, and the power spectrum of the FFT image is calculated based on the phase information;

[0086] Determine whether one or more values ​​in the power spectrum are 0. If they are not 0, then the FFT image is confirmed to contain noise.

[0087] In one embodiment, the noise determination unit 530 is further configured to:

[0088] The previous FFT image is acquired and filtered to obtain the filtered FFT image;

[0089] Determine if there is noise in the filtered FFT image. If there is noise, continue filtering the current FFT image. If there is no noise, perform an inverse Fourier transform on the current FFT image to obtain the ideal loss light image.

[0090] In one embodiment, the second processing unit 540 is used to:

[0091] The final STED image is obtained by subtracting the ideal loss light image from the initial STED image.

[0092] This invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the digital image processing method as described in the foregoing embodiments.

[0093] This invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the digital image processing method as described in the foregoing embodiments.

[0094] Those skilled in the art will readily understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0095] In the embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Units with the same function may be grouped into one unit. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices, or units, or it may be an electrical, mechanical, or other form of connection.

[0096] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention, depending on actual needs.

[0097] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0098] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), magnetic disks, or optical disks.

[0099] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A digital image processing method for improving STED imaging quality, characterized in that, include: The fluorescent beads were imaged using a STED microscope to obtain initial STED images and co-aggregation images; The initial STED image and co-aggregated image are subjected to the first image processing to obtain the actual loss light image. The actual loss light image is then subjected to Fourier transform processing and truncation processing to obtain the FFT image. The FFT image is compared with the ideal loss light FFT image, and it is determined whether there is noise in the FFT image. If there is noise, the FFT image is thresholded, and then the thresholded FFT image is subjected to inverse Fourier transform to obtain the ideal loss light image. The initial STED image and the ideal loss light image are subjected to a second image processing to obtain the final STED image; The first image processing step of the initial STED image and the co-aggregate image to obtain the actual loss light image includes: comparing the initial STED image and the co-aggregate image to obtain a comparison result; if the comparison result shows a difference, subtracting the co-aggregate image from the initial STED image to obtain the actual loss light image. The step of performing Fourier transform processing and cropping processing on the actual loss light image to obtain an FFT image includes: performing Fourier transform processing on the actual loss light image to obtain an actual loss light FFT image; and using image software to crop an image region with a predetermined amplitude value from the actual loss light FFT image to obtain an FFT image. The step of determining whether the FFT image contains noise includes: obtaining the amplitude value of the FFT image, calculating the spectrum of the FFT image based on the amplitude value; comparing the spectrum of the FFT image with the spectrum of the ideal loss light FFT image, determining whether a specific region in the spectrum of the FFT image is 0, and if it is not 0, then confirming that the FFT image contains noise. The step of determining whether the FFT image contains noise includes: acquiring the phase information of the FFT image, calculating the power spectrum of the FFT image based on the phase information; determining whether one or more values ​​in the power spectrum are 0, and if they are not 0, confirming that the FFT image contains noise. If noise exists, the FFT image is thresholded, and then the thresholded FFT image is subjected to inverse Fourier transform to obtain an ideal loss light image. This includes: acquiring the previous FFT image and filtering the FFT image to obtain a filtered FFT image; determining whether the filtered FFT image has noise; if noise exists, continuing to filter the current FFT image; if no noise exists, performing an inverse Fourier transform on the current FFT image to obtain an ideal loss light image. The step of performing a second image processing on the initial STED image and the ideal loss light image to obtain the final STED image includes: subtracting the ideal loss light image from the initial STED image to obtain the final STED image.

2. A digital image processing apparatus for improving STED imaging quality, used to implement the digital image processing method for improving STED imaging quality as described in claim 1, characterized in that, include: The imaging unit is used to image fluorescent beads using a STED microscope to obtain initial STED images and co-aggregation images. The first processing unit is used to perform a first image processing on the initial STED image and the co-aggregation image to obtain an actual loss light image, and to perform Fourier transform processing and truncation processing on the actual loss light image to obtain an FFT image. The noise determination unit is used to compare the FFT image with the ideal loss light FFT image and determine whether there is noise in the FFT image. If there is noise, the FFT image is thresholded and then the thresholded FFT image is subjected to inverse Fourier transform to obtain the ideal loss light image. The second processing unit is used to perform a second image processing on the initial STED image and the ideal loss light image to obtain the final STED image. The first processing unit is specifically used to compare the initial STED image and the co-aggregated image to obtain a comparison result; if the comparison result shows a difference, the co-aggregated image is subtracted from the initial STED image to obtain the actual loss light image; The actual loss light image is subjected to Fourier transform processing to obtain the actual loss light FFT image; the image region with a predetermined amplitude value in the actual loss light FFT image is cropped using image software to obtain the FFT image. The noise determination unit is specifically used to obtain the amplitude value of the FFT image, calculate the spectrum of the FFT image based on the amplitude value, compare the spectrum of the FFT image with the spectrum of the ideal loss light FFT image, determine whether a specific region in the spectrum of the FFT image is 0, and if it is not 0, confirm that the FFT image contains noise. The phase information of the FFT image is obtained, and the power spectrum of the FFT image is calculated based on the phase information. It is then determined whether one or more values ​​in the power spectrum are 0. If they are not 0, it is confirmed that the FFT image contains noise. The previous FFT image is acquired and filtered to obtain a filtered FFT image. It is then determined whether there is noise in the filtered FFT image. If there is noise, the current FFT image is filtered again. If there is no noise, the current FFT image is subjected to an inverse Fourier transform to obtain an ideal loss light image. The second processing unit is specifically used to subtract the ideal loss light image from the initial STED image to obtain the final STED image.

3. A computer device, characterized in that, The method includes 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 digital image processing method as described in claim 1.

4. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the digital image processing method as described in claim 1.

Images