A two-photon microscope and related microscopic imaging method

Abstract

The present application relates to the field of microscope imaging processing technology, and specifically to a two-photon microscope and related microscopic imaging method, the method comprising: obtaining tomographic images taken by a twin microscope at various focal lengths; obtaining the plane frequency of each pixel in the tomographic image according to the local grayscale change of the tomographic image; obtaining the high-frequency interpolation weight between corresponding pixels in adjacent tomographic images according to the local grayscale change of the image composed of pixels in the same row and column in all tomographic images; obtaining the interpolation control weight to weight the grayscale value of each pixel according to the spatial distribution characteristics and high-frequency interpolation weight of all edge pixels in two adjacent tomographic images, and obtaining an interpolation image between two adjacent tomographic images; obtaining the microscopic imaging result of the twin microscope according to all tomographic images and interpolation images. The present application can improve the imaging quality of a two-photon microscope.

Description

Technical Field

[0001] The present application relates to the field of microscope imaging processing technology, and in particular to a two-photon microscope and related microscopic imaging methods. Background Art

[0002] Two-photon microscopy is a fluorescence microscopy technique based on the principle of multiphoton excitation. It has three-dimensional tomography capabilities and can generate three-dimensional information images of objects by measuring the tomographic images of objects. It is widely used in the fields of medicine and cytology. When using two-photon microscopy to generate three-dimensional information images, in order to improve the longitudinal resolution of the image, it is usually necessary to interpolate between tomographic images.

[0003] Existing interpolation methods mostly use linear interpolation or high-order nonlinear interpolation, usually using smooth curves to fit known data, which is essentially a low-pass process; the boundaries of objects in the image are mainly determined by high-frequency components, and low-pass processing will ignore the information of high-frequency components, which will lead to blurred boundaries in the interpolation results. Therefore, in two-photon microscopy imaging, when interpolating between tomographic images, because the existing interpolation methods ignore the high-frequency information of the edges, the interpolated images between tomographic images have blurred edges, which in turn affects the final three-dimensional imaging effect. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a two-photon microscope and related microscopic imaging methods. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides a two-photon microscope imaging method, the method comprising the following steps:

[0006] S1, obtaining the tomographic images taken by the twin microscope at each focal length;

[0007] S2, obtaining the plane frequency of each pixel in the tomographic image according to the local grayscale change of the tomographic image; obtaining the spatial frequency of each pixel in the tomographic image according to the local grayscale change of the image composed of the pixels in the same row and column in all the tomographic images;

[0008] S3, obtaining high-frequency interpolation weights between corresponding pixel points in adjacent tomographic images according to the spatial frequency and plane frequency differences between corresponding pixel points in adjacent tomographic images;

[0009] S4, obtaining edge pixel points of each tomographic image; obtaining detail interpolation weights between corresponding pixel points in two adjacent tomographic images according to the spatial distribution characteristics of all edge pixel points in two adjacent tomographic images;

[0010] S5, comprehensively analyzing the detail interpolation weights and high-frequency interpolation weights between corresponding pixels in two adjacent tomographic images to obtain interpolation control weights; weighting the grayscale value of each pixel according to the interpolation control weights between all the corresponding pixels to obtain an interpolation image between two adjacent tomographic images; and obtaining the microscopic imaging result of the twin microscope based on all tomographic images and the interpolation images.

[0011] Furthermore, obtaining the plane frequency of each pixel point in the tomographic image according to the local grayscale change of the tomographic image includes:

[0012] For each tomographic image, a gradient image of the tomographic image is obtained, any pixel point in the gradient image is recorded as a target point, and the gray value of the target point is recorded as the plane frequency of the pixel point in the tomographic image at the same position as the target point.

[0013] Furthermore, the method for acquiring the spatial frequency of each pixel in the tomographic image includes:

[0014] The image composed of the x-th row of pixels of all tomographic images is recorded as the x-th row change image; the image composed of the y-th column of pixels of all tomographic images is recorded as the y-th column change image; for each row change image and column change image, the transverse frequency and the longitudinal frequency of each pixel point are obtained according to the method for obtaining the plane frequency of each pixel point in the tomographic image; the average value of the transverse frequency and the longitudinal frequency of each pixel point in the tomographic image is taken as the spatial frequency of each pixel point in the tomographic image.

[0015] Furthermore, the method for obtaining high-frequency interpolation weights between corresponding pixels in adjacent tomographic images includes:

[0016] For each tomographic image, the difference between the spatial frequency and the plane frequency of each pixel in the tomographic image is calculated to obtain the sum of the difference and a preset constant; the ratio of the spatial frequency of each pixel in the tomographic image to the sum is used as the high-frequency similarity of each pixel in the tomographic image; based on the high-frequency similarity of two adjacent tomographic images, the high-frequency interpolation weight between the corresponding pixels in the two adjacent tomographic images is obtained.

[0017] Furthermore, obtaining the high-frequency interpolation weights between corresponding pixels in the two adjacent tomographic images according to the high-frequency similarity of the two adjacent tomographic images includes:

[0018] For the adjacent nth and n+1th tomographic images, the sum of the high-frequency similarities of the pixels in the xth row and yth column in the nth tomographic image and the n+1th tomographic image is calculated and recorded as the first sum; the ratio of the high-frequency similarity of the pixels in the xth row and yth column in the nth tomographic image to the first sum is used as the high-frequency interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images.

[0019] Furthermore, the method for obtaining the detail interpolation weight between corresponding pixels in two adjacent tomographic images includes: recording the detail interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images as , ;in, is a linear normalization function, , are the number of edge pixels of the nth and n+1th tomographic images, respectively. , are the Euclidean distances between the pixel point in the xth row and yth column and the kth and ith edge pixel points in the nth and n+1th tomographic images, respectively.

[0020] Furthermore, the method for obtaining the interpolation control weight includes: recording the interpolation control weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images as , ; In the formula, is the high-frequency interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images, P is the preset mean interpolation weight, is the detail interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images.

[0021] Furthermore, the method for obtaining the interpolation image between two adjacent tomographic images includes: recording the grayscale value of the pixel point in the xth row and yth column in the interpolation image between the nth and n+1th tomographic images as , ; In the formula, is the interpolation control weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images, , are the grayscale values ​​of the pixel at the xth row and yth column in the nth and n+1th tomographic images respectively.

[0022] Furthermore, the microscopic imaging results of the twin microscope are obtained according to all the tomographic images and the interpolated images, specifically including:

[0023] A three-dimensional reconstruction algorithm is used to calculate all interpolation images and tomographic images to obtain a microscopic imaging result of a twin microscope; the microscopic imaging result of the twin microscope is a three-dimensional model.

[0024] In a second aspect, an embodiment of the present application also provides a two-photon microscope, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of a two-photon microscope imaging method described in any one of the above items are implemented.

[0025] This application has at least the following beneficial effects:

[0026] This application analyzes the edge blur problem of interpolation images between tomographic images in the process of two-photon microscopy. First, the plane frequency of the tomographic image is calculated, and then the tomographic images are arranged to obtain the gray gradient change of the tomographic image formed by continuously adjusting the focal length in three-dimensional space; further, the plane frequency and the spatial frequency are compared to screen out the non-noise high-frequency information part, and the high-frequency interpolation weight is obtained to characterize the high-frequency information position of the three-dimensional object in the longitudinal direction; further, through the texture information of the tomographic image, the position with high texture density is recorded as the detail position, and the detail interpolation weight is calculated to judge the importance of the high-frequency information; the interpolation control weight is obtained by combining the high-frequency interpolation weight and the detail interpolation weight, and the mean interpolation is used for the smooth part of the three-dimensional object, and the high-frequency weight is used for interpolation for the detail part, and the detail interpolation weight is used to weigh between the two interpolation methods; the problem of rough edges of three-dimensional objects caused by ignoring high-frequency information in the traditional interpolation algorithm is avoided. Finally, the three-dimensional information reconstruction of the tomographic image is completed according to the obtained interpolation image, the microscopic imaging of the two-photon microscope is completed, and the imaging quality is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0028] Figure 1 A flowchart of a two-photon microscope imaging method provided in one embodiment of the present application;

[0029] Figure 2 A block diagram for obtaining spatial frequency provided in accordance with one embodiment of the present application. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0031] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0032] The specific scheme of a two-photon microscope and related microscopic imaging method provided by the present application is described in detail below with reference to the accompanying drawings.

[0033] See also Figure 1 , which shows a flowchart of a two-photon microscope imaging method provided by an embodiment of the present application, the method comprising the following steps:

[0034] S1, obtain the tomographic images taken by the twin microscope at various focal lengths.

[0035] A specific implementation scenario of the embodiment of the present invention is the interpolation image generation scenario in the twin microscope imaging process; when the tomographic images of the twin microscope are used to reconstruct a three-dimensional model, the interpolation image usually ignores the information of the high-frequency component due to low-pass processing, which leads to the problem of blurred boundaries in the interpolation result. First, N tomographic images are obtained by continuously increasing the focal length of the twin microscope. That is, each tomographic image corresponds to a focal length, and the tomographic image is a grayscale image.

[0036] The main reason for the blurred boundaries of the interpolation results is that the traditional interpolation algorithm calculates the interpolation results by fitting the original data with a smooth curve, ignoring the frequency domain information of the image space. The boundaries of objects in the image are mainly determined by the high-frequency components. Therefore, this embodiment extracts the spatial frequency domain information between consecutive tomographic images to interpolate the tomographic images.

[0037] S2, obtaining the plane frequency of each pixel in the tomographic image according to the local grayscale change of the tomographic image; obtaining the spatial frequency of each pixel in the tomographic image according to the local grayscale change of the image composed of the pixels in the same row and column in all tomographic images.

[0038] This embodiment performs tomographic image interpolation based on frequency information. The frequency of an image represents the intensity of grayscale changes in the image and is the gradient of grayscale in a plane space. The high frequency of an image indicates an intense grayscale change; the low frequency indicates a gentle grayscale change. In order to fully extract the spatial frequency characteristics of all tomographic images, interpolated images are generated using data with higher frequencies. Tomographic images formed with different focal lengths represent different imaging positions. This embodiment extracts high-frequency information through the spatial frequency information of tomographic images with continuously changing positions.

[0039] First, for each tomographic image, a gradient image of the tomographic image is obtained, any pixel point in the gradient image is recorded as a target point, and the gray value of the target point is recorded as the plane frequency of the pixel point in the tomographic image at the same position as the target point.

[0040] Specifically, this embodiment takes the nth tomographic image as an example, adopts the Sobel operator for calculation, and outputs a gradient image of the nth tomographic image. The size of the gradient image is the same as that of the tomographic image, and the grayscale value of the pixel point in the xth row and yth column of the gradient image represents the plane frequency of the pixel point in the xth row and yth column of the nth tomographic image.

[0041] The frequency information obtained from the gradient image is the planar grayscale variation characteristics of the tomographic image. In order to further analyze the grayscale variation characteristics at different focal lengths, this embodiment extracts the frequency information from a direction perpendicular to the plane where the tomographic image is located.

[0042] Specifically, the image composed of the x-th row of pixels of all tomographic images is recorded as the x-th row change image; the image composed of the y-th column of pixels of all tomographic images is recorded as the y-th column change image; for each row change image and column change image, the transverse frequency and the longitudinal frequency of each pixel in the tomographic image are obtained according to the method for obtaining the plane frequency of each pixel in the tomographic image; the average value of the transverse frequency and the longitudinal frequency of each pixel in the tomographic image is taken as the spatial frequency of each pixel in the tomographic image.

[0043] In this embodiment, the x-th row change image is taken as input, the Sobel operator is used for calculation, and the output is recorded as the gradient image of the x-th row change image, and the grayscale value of each pixel in the gradient image is recorded as the lateral frequency of the corresponding pixel in the n-th tomographic image; similarly, the y-th column change image is taken as input, the Sobel operator is used for calculation, and the output is recorded as the gradient image of the y-th column change image, and the grayscale value of each pixel in the gradient image is recorded as the longitudinal frequency of the corresponding pixel in the n-th tomographic image.

[0044] The spatial frequency calculated by integrating the transverse frequency and the longitudinal frequency represents the grayscale gradient change of the tomographic image formed by continuously adjusting the focal length in the three-dimensional space.

[0045] The spatial frequency acquisition block diagram is as follows: Figure 2 shown.

[0046] S3, obtaining high-frequency interpolation weights between corresponding pixel points in adjacent tomographic images according to the spatial frequency and plane frequency differences between corresponding pixel points in adjacent tomographic images.

[0047] In order to solve the problem of blurred edges of interpolated images when traditional interpolation algorithms are used, it is necessary to retain the high-frequency information in the tomographic image. Since a new interpolated image is generated and inserted between two tomographic images during interpolation, when two tomographic images are used to generate an intermediate interpolated image, the image information of the tomographic image with a high spatial frequency tends to be selected to generate the intermediate interpolated image.

[0048] At the same time, since white noise may exist in the tomographic image, and white noise is usually a higher frequency part, generating an intermediate interpolated image only with a tomographic image with a high spatial frequency is likely to cause more noise in the interpolated image, affecting the image quality. Since the two-photon microscope obtains three-dimensional information, if the difference between the plane frequency and the spatial frequency of the object is small, the frequency information of the plane frequency is more likely to be the frequency information of the object itself, rather than the frequency information of the noise, so the noise point can be identified by comparing the spatial frequency and the plane frequency.

[0049] Specifically, for each tomographic image, the difference between the spatial frequency and the plane frequency of each pixel in the tomographic image is calculated to obtain the sum of the difference and a preset constant; the ratio of the spatial frequency of each pixel in the tomographic image to the sum is used as the high-frequency similarity of each pixel in the tomographic image.

[0050] In this embodiment, the high-frequency similarity of the pixel point in the xth row and yth column in the nth tomographic image is The calculation formula is:

[0051] ;

[0052] In the formula, is the high-frequency similarity of the pixel point in the xth row and yth column in the nth tomographic image, , are respectively the spatial frequency and plane frequency of the pixel point in the xth row and yth column in the nth tomogram, and s is a preset constant greater than 0 in order to prevent the denominator from being 0. In this embodiment, it is 1.

[0053] The smaller the difference between the plane frequency and the spatial frequency, the more uniform the frequencies of the corresponding pixels in different directions in the three-dimensional space, and the less likely their frequency characteristics are white noise characteristics during shooting. At the same time, the larger the spatial frequency of the xth row and yth column in the nth tomogram, the larger its frequency in the vertical direction, and the more the image information of the nth tomogram should be used when interpolating using the nth tomogram.

[0054] Furthermore, according to the high-frequency similarity of two adjacent tomographic images, a high-frequency interpolation weight between corresponding pixels in the two adjacent tomographic images is obtained; for the adjacent nth and n+1th tomographic images, the sum of the high-frequency similarities of the pixels in the xth row and yth column in the nth tomographic image and the n+1th tomographic image is calculated and recorded as the first sum; the ratio of the high-frequency similarity of the pixels in the xth row and yth column in the nth tomographic image to the first sum is used as the high-frequency interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images.

[0055] In this embodiment, the high-frequency interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images is recorded as , ;in, , are the high-frequency similarities of the pixels in the xth row and yth column in the nth tomogram and the n+1th tomogram, respectively.

[0056] The higher the high-frequency interpolation weight of the pixels between the nth and n+1th tomographic images, the higher the weight of the nth tomographic image when the nth tomographic image generates the nth interpolation image, and the higher the corresponding high-frequency interpolation weight.

[0057] S4, obtaining edge pixel points of each tomographic image; and obtaining detail interpolation weights between corresponding pixel points in two adjacent tomographic images according to spatial distribution characteristics of all edge pixel points in two adjacent tomographic images.

[0058] When performing image interpolation, the traditional method uses the idea of ​​low-pass processing to interpolate, the purpose of which is to ensure that the texture features of the interpolated image are consistent with the original image, that is, to ensure that the grayscale value at the interpolated position changes smoothly.

[0059] If high-frequency image information is used for interpolation, the smoothness of the interpolated image may be affected for image positions with less texture information in the tomographic image, making the final three-dimensional imaging result rough. Therefore, this embodiment uses high-frequency similarity to assist in generating interpolated images for positions with more texture information in the tomographic image, and uses traditional methods to generate interpolated images for positions with less texture information, thereby avoiding the problem of rough final three-dimensional imaging.

[0060] Specifically, the detail interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images is recorded as , ;in, is a linear normalization function, , are the number of edge pixels of the nth and n+1th tomographic images, respectively. , are the Euclidean distances between the pixel point in the xth row and yth column and the kth and ith edge pixel points in the nth and n+1th tomographic images, respectively.

[0061] In this embodiment, the Canny edge detection algorithm is used to calculate and obtain the binary image of each tomographic image. The gray value of 0 in the image is the pixel position where no edge texture is detected, and the gray value of 1 is the edge pixel position where the edge texture is detected. In the formula, , is the number of tomographic images.

[0062] Among them, the more edge pixels there are near the xth row and yth column in the nth tomogram, the higher the texture density, which means that the pixel position is a more complex detail part in the three-dimensional image, and the corresponding interpolation image of the tomogram should be generated using high-frequency similarity, so as to solve the problem of interpolation image blur and avoid the problem of rough final three-dimensional imaging.

[0063] S5, comprehensively analyzing the detail interpolation weights and high-frequency interpolation weights between corresponding pixels in two adjacent tomographic images to obtain interpolation control weights; weighting the grayscale value of each pixel according to the interpolation control weights between all the corresponding pixels to obtain an interpolation image between two adjacent tomographic images; and obtaining the microscopic imaging result of the twin microscope based on all tomographic images and the interpolation images.

[0064] According to the above analysis, when generating an interpolated image from a tomographic image, the retention of the high-frequency information of the tomographic image and the detailed edge information of the image both have a significant improvement effect on the edge blur problem of the interpolated image. Therefore, the interpolated image is generated by comprehensively considering the high-frequency information and the detailed edge information of the image.

[0065] Specifically, the interpolation control weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images is recorded as , ; In the formula, is the high-frequency interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images, and P is the preset mean interpolation weight. In the traditional mean interpolation algorithm, , is the detail interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images.

[0066] Among them, the high-frequency interpolation weight represents the consideration ratio of high and low frequencies when generating interpolated grayscale values; It is the interpolation weight used when using mean interpolation, which means that the same weight is used for all gray values ​​to generate the interpolated image; the detail interpolation weight represents the trade-off between the two interpolation methods. The smaller the detail interpolation weight, the more likely it is that the mean interpolation method will be used to generate the interpolated image; the larger the detail interpolation weight, the more high-frequency information should be considered.

[0067] Finally, according to the interpolation control weights between all corresponding pixels in two adjacent tomographic images, the interpolation image between two adjacent tomographic images is obtained: the gray value of the pixel in the xth row and yth column in the interpolation image between the nth and n+1th tomographic images is recorded as , ; In the formula, is the interpolation control weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images, , are the grayscale values ​​of the pixel at the xth row and yth column in the nth and n+1th tomographic images respectively.

[0068] The calculation of the interpolation image is completed with the interpolation control weight as the weight. The larger the high-frequency interpolation weight, the greater the weight of the nth tomogram compared to the n+1th tomogram, and the greater the weight of the nth tomogram when calculating the interpolation grayscale. The interpolation control weight consists of two parts: high-frequency similarity and detail interpolation weight. The high-frequency similarity ensures that high-frequency information is used as the main information when calculating the interpolation image, avoiding the edge blur problem caused by the traditional interpolation algorithm. The detail interpolation weight makes it possible to use high-frequency information to calculate the interpolation grayscale value when calculating the interpolation image for the detail part, avoiding the problem of image non-smoothing caused by using high-frequency information for image interpolation.

[0069] Finally, all interpolation images and tomographic images are input, and the three-dimensional reconstruction algorithm of the tomographic images is used for calculation. This embodiment adopts a filtered back-projection reconstruction algorithm, and the output is a reconstructed three-dimensional model to complete the microscopic imaging of the twin microscope.

[0070] Since this embodiment utilizes the tomographic image itself and the high-frequency information between the tomographic images, the traditional interpolation method is improved, the problem of blurred edges of the interpolated image caused by the traditional method ignoring high-frequency information is solved, and the final two-photon microscope microscopic imaging quality is improved.

[0071] Based on the same inventive concept as the above method, an embodiment of the present application also provides a two-photon microscope, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of any one of the above-mentioned two-photon microscope imaging methods are implemented.

[0072] Through the above description of the implementation method in combination with the accompanying drawings, technical personnel in the relevant field can understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0073] The above contents are only specific implementation methods of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be covered by the protection scope of the present application.

Claims

1. A two-photon microscope imaging method, characterized in that: The method comprises the following steps: S1, obtaining the tomographic images taken by the twin microscope at each focal length; S2, obtaining the plane frequency of each pixel in the tomographic image according to the local grayscale change of the tomographic image; obtaining the spatial frequency of each pixel in the tomographic image according to the local grayscale change of the image composed of the pixels in the same row and column in all the tomographic images; S3, obtaining high-frequency interpolation weights between corresponding pixel points in adjacent tomographic images according to the spatial frequency and plane frequency differences between corresponding pixel points in adjacent tomographic images; S4, obtaining edge pixel points of each tomographic image; obtaining detail interpolation weights between corresponding pixel points in two adjacent tomographic images according to the spatial distribution characteristics of all edge pixel points in two adjacent tomographic images; S5, combining the detail interpolation weights and high-frequency interpolation weights between corresponding pixels in two adjacent tomographic images to obtain an interpolation control weight; weighting the grayscale value of each pixel according to the interpolation control weights between all the corresponding pixels to obtain an interpolation image between two adjacent tomographic images; and obtaining the microscopic imaging result of the twin microscope according to all tomographic images and the interpolation image; The method for obtaining high-frequency interpolation weights between corresponding pixel points in adjacent tomographic images includes: For each tomographic image, the difference between the spatial frequency and the plane frequency of each pixel in the tomographic image is calculated to obtain the sum of the difference and a preset constant; the ratio of the spatial frequency of each pixel in the tomographic image to the sum is used as the high-frequency similarity of each pixel in the tomographic image; and the high-frequency interpolation weight between corresponding pixels in the two adjacent tomographic images is obtained according to the high-frequency similarity of the two adjacent tomographic images; The method for obtaining the detail interpolation weight between corresponding pixel points in two adjacent tomographic images comprises: recording the detail interpolation weight between the pixel points in the xth row and yth column in the nth and n+1th tomographic images as , ;in, is a linear normalization function, , are the number of edge pixels of the nth and n+1th tomographic images, respectively. , are the Euclidean distances between the pixel point in the xth row and yth column and the kth and ith edge pixel points in the nth and n+1th tomographic images, respectively.

2. A two-photon microscope imaging method as claimed in claim 1, characterized in that: The method of obtaining the plane frequency of each pixel point in the tomographic image according to the local grayscale change of the tomographic image includes: For each tomographic image, a gradient image of the tomographic image is obtained, any pixel point in the gradient image is recorded as a target point, and the gray value of the target point is recorded as the plane frequency of the pixel point in the tomographic image at the same position as the target point.

3. A two-photon microscope imaging method as claimed in claim 2, characterized in that: The method for acquiring the spatial frequency of each pixel point in the tomographic image comprises: The image composed of the x-th row of pixels of all tomographic images is recorded as the x-th row change image; the image composed of the y-th column of pixels of all tomographic images is recorded as the y-th column change image; for each row change image and column change image, the transverse frequency and the longitudinal frequency of each pixel point are obtained according to the method for obtaining the plane frequency of each pixel point in the tomographic image; the average value of the transverse frequency and the longitudinal frequency of each pixel point in the tomographic image is taken as the spatial frequency of each pixel point in the tomographic image.

4. A two-photon microscope imaging method as claimed in claim 1, characterized in that: The step of obtaining the high-frequency interpolation weight between corresponding pixels in the two adjacent tomographic images according to the high-frequency similarity of the two adjacent tomographic images comprises: For the adjacent nth and n+1th tomographic images, the sum of the high-frequency similarities of the pixels in the xth row and yth column in the nth tomographic image and the n+1th tomographic image is calculated and recorded as the first sum; the ratio of the high-frequency similarity of the pixels in the xth row and yth column in the nth tomographic image to the first sum is used as the high-frequency interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images.

5. A two-photon microscope imaging method as claimed in claim 1, characterized in that: The method for obtaining the interpolation control weight comprises: recording the interpolation control weight between the pixel points in the xth row and yth column in the nth and n+1th tomographic images as , ; In the formula, is the high-frequency interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images, P is the preset mean interpolation weight, is the detail interpolation weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images.

6. A two-photon microscope imaging method as claimed in claim 1, characterized in that: The method for obtaining an interpolation image between two adjacent tomographic images comprises: recording the gray value of the pixel point in the xth row and yth column in the interpolation image between the nth and n+1th tomographic images as , ; In the formula, is the interpolation control weight between the pixels in the xth row and yth column in the nth and n+1th tomographic images, , are the grayscale values ​​of the pixel at the xth row and yth column in the nth and n+1th tomographic images respectively.

7. A two-photon microscope imaging method as claimed in claim 1, characterized in that: The microscopic imaging results of the twin microscope are obtained according to all the tomographic images and the interpolation images, specifically including: A three-dimensional reconstruction algorithm is used to calculate all interpolation images and tomographic images to obtain a microscopic imaging result of a twin microscope; the microscopic imaging result of the twin microscope is a three-dimensional model.

8. A two-photon microscope comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of a two-photon microscope imaging method as described in any one of claims 1-7 are implemented.

Images