Automatic registration and spatial alignment method and system based on multi-channel fluorescence image, electronic equipment and storage medium

Abstract

The invention discloses an automatic registration and spatial alignment method and system based on a multi-channel fluorescence image, electronic equipment and a storage medium, and the method comprises the steps: setting a first channel as a focusing channel, obtaining n focuses for focusing, and collecting a focusing image; comparing the gray histograms of the n images, selecting two images with obvious wave crests and horizontal focal positions, and recording the focusing positions of the images; equally dividing the selected image into block images, and traversing and selecting two areas with most obvious wave crests of a histogram; focusing and image acquisition of a next channel image are carried out to obtain a clear grayscale image; according to the region, carrying out feature value matching on the image at the corresponding position to obtain an offset coordinate of the clear grayscale image; obtaining the offset coordinate of the next channel based on the offset coordinate; the offset coordinates of all channels are obtained; and performing registration and superposition on the complete images of all the channels according to the obtained offset to form a multi-channel fluorescence image. And the splicing accuracy of single channels is ensured, and the time cost is saved.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and in particular, to an automatic registration and spatial alignment method, system, electronic device, and storage medium based on multi-channel fluorescence images. Background Art

[0002] In the fields of tumor microenvironment research, immunology, pathological diagnosis, etc., multiplex immunofluorescence (mIF) technology has become an important tool due to its advantages such as multiplex detection, high resolution, and spatial information retention. In the process of converting multi-channel fluorescence slices into digital images, a multi-channel fluorescence digital slice scanner is required. The fluorescence slice sample is fixed on the platform of the scanner, and the scanning platform realizes precise movement and positioning of the slice by controlling the translation of the tray in the XYZ three axes, so that the image acquisition is completed through the optical system above the scanning platform. By irradiating the slice with light sources of different wavelengths, the fluorescence signals of different channels can be excited, and thus the slice can be scanned channel by channel to obtain multi-channel fluorescence images.

[0003] However, during the scanning process, due to multiple movements of the platform, its cumulative error and the rapid movement of the slice may cause inaccurate matching or inaccurate splicing of different channels of the multi-channel fluorescence image. Different channels of the fluorescent tissue correspond to specific molecules or structures. If the matching is inaccurate, it will cause loss of image information, thereby affecting the accuracy of medical diagnosis and pathological research. In the prior art, since the image content scanned by each channel represents different biological proteins or molecular structures, the image information at the same position in each layer is inconsistent, so it is impossible to perform matching through conventional image algorithms. Currently, the technology basically switches different channels at the same position to eliminate the error of slice movement, which greatly increases the scanning time. Summary of the Invention

[0004] The purpose of the present application is to provide an automatic registration and spatial alignment method, system, electronic device, and storage medium based on multi-channel fluorescence images to solve one or more technical problems existing in the prior art, and at least provide a beneficial choice or create conditions.

[0005] The present application adopts the following technical solutions to achieve the above-mentioned invention purpose: The present application provides an automatic registration and spatial alignment method based on multi-channel fluorescence images, including: S1, obtaining n focal points for focusing, setting the first channel as the focusing channel, and saving the clear grayscale focusing images collected during focusing; S2, comparing the clear grayscale focusing images and , ..., The grayscale histograms of are obtained, and two focus images with obvious peaks and horizontal focus positions are selected. and , and record the focus positions of the focus images and ; and ; S3, divide and into block diagrams, and traverse to select the regions with the most obvious two peaks in the histogram and ; and ; S4, then use the focus channel to focus at the focus positions and and collect focus images to obtain clear grayscale focus maps , clear grayscale focus map , clear grayscale focus map and clear grayscale focus map ; S5, according to the regions and , perform eigenvalue matching on the images at the corresponding positions to obtain the offset coordinates of the clear grayscale focus maps , clear grayscale focus map , clear grayscale focus map and clear grayscale focus map ; S6, based on the offset coordinates of the clear grayscale focus maps , clear grayscale focus map , clear grayscale focus map and clear grayscale focus map , obtain the offset coordinates of the next channel relative to the first channel; S7, repeat S4 - S6 until the offset coordinates of all channels relative to the first channel and the complete images of all channels are obtained; S8, register and stack the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image.

[0006] Furthermore, obtain foci for focusing, set the first channel as the focus channel, and save the clear grayscale focus images collected during focusing, specifically including: Obtain the foci selected on the software operation interface of the fluorescence scanner , , ..., ; Perform image focusing with the first channel as the focusing channel, and perform focusing at the focal position in sequence; Save the clear grayscale images , ,..., .

[0007] Furthermore, compare the clear grayscale focus images , ,..., histograms of gray levels, and select two focus images with obvious peaks and the focal positions being horizontal and , record the focal positions and of the focus images and , specifically including: Compare the focus images , ,..., histograms of gray levels; Select two of the focus images, whose histograms both show two obvious and separated peaks, the valleys are deep and clear, and the coordinates of the focal positions corresponding to the two images are the same in the X-axis or Y-axis direction, and record the focal positions and of the focus images and .

[0008] Furthermore, divide and into block images equally, traverse and select the regions and where the two peaks of the histogram are the most obvious, specifically including: Divide into 3x3 block images equally, traverse the block images and generate the histograms of gray levels for each block image, and select the regions where the two peaks of the histogram are the most obvious; Divide into 3x3 block images equally, traverse the block images and generate the histograms of gray levels for each block image, and select the regions where the two peaks of the histogram are the most obvious.

[0009] Furthermore, use the focusing channel to perform focusing at the focal positions and and collect the focus images, obtaining clear grayscale focus images , clear grayscale focus images , clear grayscale focus images and clear grayscale focus images , specifically including: Use the focus channel for focusing and move the platform to the position, from the position to the position for focusing and save the captured clear grayscale focus image ; From the position to the position for focusing and save the captured clear grayscale focus image ; From the position to the position for focusing and save the captured clear grayscale focus image ; From the position to the position for focusing and save the captured clear grayscale focus image ; Among them and are the images of the next channel corresponding to the first channel position, and are the images of the next channel corresponding to the first channel position.

[0010] Furthermore, according to the regions and , perform eigenvalue matching on the images at the corresponding positions to obtain the offset coordinates of the clear grayscale focus image , clear grayscale focus image , clear grayscale focus image and clear grayscale focus image , specifically including: Perform eigenvalue matching on the regions and of the images to obtain the offset coordinate ; Perform eigenvalue matching on the regions and of the images to obtain the offset coordinate ; Perform eigenvalue matching on the regions and of the images to obtain the offset coordinate ; Perform eigenvalue matching on the regions and of the images Perform eigenvalue matching to obtain the offset coordinates .

[0011] Further, based on the clear grayscale focus maps , clear grayscale focus maps , clear grayscale focus maps and clear grayscale focus maps of the offset coordinates, obtain the offset coordinates of the next channel relative to the first channel, specifically including: Calculate the average value of the four offset coordinates and the standard deviation of each offset coordinate , , , , the formula is as follows: ; After removing the coordinate with the largest deviation among the four offset coordinates, calculate the average value, and the formula is as follows: ; This average value is the offset of the overall image of the next channel relative to the overall image of the first channel in the x - direction and y - direction.

[0012] This application provides an automatic registration and spatial alignment system based on multi - channel fluorescence images, including: An acquisition unit for acquiring focuses for focusing, setting the first channel as the focus channel, and saving the clear grayscale focus images collected during focusing; A comparison unit for comparing the clear grayscale focus images , ,..., of the grayscale histograms, selecting two focus images with obvious peaks and the focus positions being horizontal and , recording the focus positions and of the focus images and ; A selection unit for equally dividing and into block images, and traversing to select the regions and with the two most obvious peaks in the histogram; A focus acquisition unit for refocusing and acquiring focus images at the focus positions and using the focus channel, and obtaining the clear grayscale focus map , Clear grayscale focus map , Clear grayscale focus map and clear grayscale focus map ; Eigenvalue matching unit, for matching the eigenvalues of the images at corresponding positions according to region and , and obtaining the offset coordinates of the clear grayscale focus map , clear grayscale focus map , clear grayscale focus map and clear grayscale focus map ; Calculation unit, for obtaining the offset coordinates of the next channel relative to the first channel based on the offset coordinates of the clear grayscale focus map , clear grayscale focus map , clear grayscale focus map , clear grayscale focus map ; Repeating unit, for repeatedly using the focus acquisition unit, eigenvalue matching unit and calculation unit until obtaining the offset coordinates of all channels relative to the first channel and the complete images of all channels; Output unit, for registering and superimposing the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image.

[0013] This application provides an electronic device, including a memory and a processor; The memory is used for storing instructions; The processor is used for operating according to the instructions to execute the steps of the foregoing automatic registration and spatial alignment method based on multi-channel fluorescence images.

[0014] This application provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the foregoing automatic registration and spatial alignment method based on multi-channel fluorescence images are implemented.

[0015] The beneficial effects of this application are as follows: This application scans the images channel by channel. After all the small images of each channel are scanned and stitched into a large image, the scanning of the next channel is carried out, and finally the images of different channels are registered and superimposed. This method not only simplifies the difficulty of stitching, ensures the accuracy of stitching of a single channel, but also eliminates the errors between different channels through the registration algorithm; different channels only need to be switched once, greatly saving the time cost. Description of the Drawings

[0016] Figure 1 It is an overall flowchart of an automatic registration and spatial alignment method based on multi-channel fluorescence images provided according to an embodiment of this application; Figure 2 A fluorescence multi-channel schematic diagram of an automatic registration and spatial alignment method based on multi-channel fluorescence images provided according to an embodiment of the present application; Figure 3 A schematic diagram of image overlap of different channels in the ideal state and the actual state of an automatic registration and spatial alignment method based on multi-channel fluorescence images provided according to an embodiment of the present application; Figure 4 A gray-scale distribution diagram of an image that meets the selection requirements of an automatic registration and spatial alignment method based on multi-channel fluorescence images provided according to an embodiment of the present application. Detailed implementation manners

[0017] During the scanning process of fluorescence images, it is necessary to use lights of different wavelengths to excite and scan the same position of the same slice multiple times. One wavelength represents one channel, and the overall image of each channel is composed of multiple small images stitched together. The scanning platform, the optical system, and the camera acquisition system work together to complete the image acquisition. During the image acquisition process, the movement of the platform will inevitably cause offsets. When the offset errors accumulate to a certain extent, they will affect the quality of the generated fluorescence images. In the prior art, at the same position, different channels are used alternately to scan the small images in turn to reduce the offsets between channels, which can eliminate the errors between single small images of different channels at the same position. For software, this method accumulates the errors into the overall image of the same channel, requiring a high-precision stitching algorithm and causing difficulties in stitching; for hardware, the frequent switching between different channels will increase the scanning time cost and is not conducive to the realization of rapid scanning of fluorescence slices. Based on this, the present invention scans the images channel by channel. After all the small images of each channel are scanned, they are stitched into a large image, then the next channel is scanned, and finally the images of different channels are registered and superimposed. For software, this method simplifies the stitching difficulty, ensures the accuracy of stitching of a single channel, and eliminates the errors between different channels through the registration algorithm; for hardware, different channels only need to be switched once, greatly saving the time cost.

[0018] As Figures 1 to 4 shown, an automatic registration and spatial alignment method based on multi-channel fluorescence images provided by the present application specifically includes the following steps: S1: Obtain focal points for focusing, set the first channel as the focusing channel, and save the clear gray-scale focusing images collected during focusing; Specifically, S1 is: Set the first channel as the focusing channel to ensure that the focusing images collected in subsequent multiple focusings have the same source. Select focal points , ,..., , perform focusing at the focal position in sequence, and save the clear grayscale focus images , ,..., . Then complete the acquisition of the channel image.

[0019] S2. Compare the clear grayscale focus images , ,..., for their grayscale histograms, and select two focus images with obvious peaks and horizontal focal positions and , and record the focal positions and of the focus images; and ; Specifically, S2 is as follows: Compare the images , ,..., for their grayscale histograms, select two images whose histograms both show two obvious and separated peaks, deep and clear valleys, and the focal positions corresponding to the two images have the same coordinates in the X-axis or Y-axis direction, and record the focal positions and of the focus images; and . Among them, the two obvious and separated peaks presented by the histogram are the foreground and background regions respectively. The foreground and background are easy to separate, and the image can better calculate the feature points; the focal positions corresponding to the two images have the same coordinates in the X-axis or Y-axis direction in order to move only in one direction when moving the focusing platform, reducing the movement error in the other direction.

[0020] S3: Divide and into block images, traverse and select the regions with the most obvious two peaks in the histogram and ; Specifically, S3 is as follows: Divide into 3x3 block images, traverse the block images and generate the grayscale histogram of each block image, and select the region with the most obvious two peaks in the histogram ; Divide into 3x3 block images, traverse the block images and generate the grayscale histogram of each block image, and select the region with the most obvious two peaks in the histogram . Among them, compared with the whole image, the block image information is simpler, easier to find feature points, and the registration algorithm has a better effect.

[0021] S4: Use the focusing channel again at the focusing position and perform focusing and collect the focused images to obtain clear grayscale focused images 、clear grayscale focused images 、clear grayscale focused images and clear grayscale focused images ; Specifically, S4 is as follows: Perform re - focusing, use the focusing channel to perform focusing and collect the focused images, move the platform to position, move from position to position to perform focusing and save the collected clear grayscale focused images , move from position to position to perform focusing and save the collected clear grayscale focused images , move from position to position to perform focusing and save the collected clear grayscale focused images , move from position to position to perform focusing and save the collected clear grayscale focused images . Among them and are the images of the next channel corresponding to the first channel position, and are the images of the next channel corresponding to the first channel position. Among them, moving from one focus position to another and then performing focusing is to simulate the movement of the platform during the normal focusing process and control other environmental factors for collecting images at the same focus each time.

[0022] S5: According to regions and , perform eigenvalue matching on the images at the corresponding positions to obtain the offset coordinates of clear grayscale focused images 、clear grayscale focused images 、clear grayscale focused images and clear grayscale focused images ; Specifically, S5 is as follows: Perform eigenvalue matching on the regions and of the focused images to obtain the offset coordinate , perform eigenvalue matching on the regions of the focused images and of the focused images Perform eigenvalue matching to obtain the offset coordinates , the focused image and of the region Perform eigenvalue matching to obtain the offset coordinates , the focused image and of the region Perform eigenvalue matching to obtain the offset coordinates . Specifically, the Scale-Invariant Feature Transform (SIFT) matching algorithm is used. The core idea of the Scale-Invariant Feature Transform (SIFT) algorithm is to achieve robustness to image transformations through multi-scale spatial analysis and local feature description. The algorithm first constructs a Gaussian pyramid scale space, detects key points at different scales, locates stable extreme points using the Difference of Gaussian (DoG) function, and filters out low-contrast and edge response points through curvature analysis to improve feature stability. Subsequently, the algorithm assigns a main direction to each key point and generates a 128-dimensional feature description vector with rotational invariance based on the gradient direction statistics within the neighborhood of the key point. This vector is formed by dividing the 16×16 neighborhood into 4×4 sub-regions and calculating the 8-direction gradient histograms in each sub-region and then concatenating them. In the feature matching stage, a strategy of nearest neighbor search combined with a ratio test is adopted. Reliable matching point pairs are found by comparing the Euclidean distances between feature vectors, and the Random Sample Consensus (RANSAC) algorithm is used to eliminate false matches to further improve the matching accuracy. Finally, through the coordinate transformation analysis of the correct matching point pairs, the offset coordinates between the images can be accurately calculated.

[0023] S6: Based on the clear grayscale focused images , the clear grayscale focused images , the clear grayscale focused images , and the clear grayscale focused images and the offset coordinates, obtain the offset coordinates of the next channel relative to the first channel; The specific content of S6 is as follows: Calculate the average value of the four offset coordinates and the standard deviation of each offset coordinate , , , , where: ; To reduce errors and ensure the accuracy of the results, remove the coordinate with the largest deviation among the four offset coordinates and then calculate the average value. The offset coordinate with the highest standard deviation is an outlier that deviates more from the mean. Take the average value of the remaining three offset coordinates to obtain , where: ; This average value is the offset of the overall image of the next channel relative to the overall image of the first channel in the x and y directions. Specifically, calculating the dynamic standard deviation and removing outliers can effectively improve the robustness of the eigenvalue matching algorithm.

[0024] S7. Repeat S4 - S6 until the offset coordinates of all channels relative to the first channel and the complete images of all channels are obtained; Specifically, S7 is as follows: If the image of the next channel has not been completely acquired, repeat steps S4 - S6 until the offsets of all channels are calculated.

[0025] S8. Register and stack the images of all channels according to the obtained offset coordinates to form a multi - channel fluorescence image; Specifically, S8 is as follows: Taking the first channel (usually DAPI nuclear staining or the clearest channel) as the reference coordinate system, a unified image space reference system is established. Other channels have preset specific pseudo - color mappings before image acquisition, and these color selections strictly correspond to the emission wavelength characteristics of specific fluorescent dyes during imaging: for example, usually the excitation wavelength of DAPI nuclear staining is 340nm, the emission wavelength is 488nm, and the imaging is blue; the excitation wavelength of FITC is 495nm, the emission wavelength is 518nm, and the imaging is green; the excitation wavelength of Cy3 is 550nm, the emission wavelength is 570nm, and the imaging is orange - red; the excitation wavelength of Cy5 is 650nm, the emission wavelength is 670nm, and the imaging is dark red, etc. After completing the spatial registration of each channel, the images of each channel are synthesized according to the preset wavelength - color mapping relationship, and finally a multi - channel fluorescence composite image with clear spectral characteristics is generated.

[0026] This application also provides an automatic registration and spatial alignment system based on a multi - channel fluorescence image, including: An acquisition unit for acquiring a number of foci for focusing, setting the first channel as the focusing channel, and saving the number of clear grayscale focusing images acquired during focusing; A comparison unit for comparing the number of clear grayscale focusing images , ,..., of the grayscale histograms, selecting two focusing images with obvious peaks and the focus positions being horizontal and , recording the focusing positions and of the focusing images and ; A selection unit, configured to and be equally divided into block diagrams, and traverse to select the regions with the most obvious two peaks of the histogram and ; A focus acquisition unit, configured to use the focus channel again at the focus position and to perform focusing and acquire a focused image, so as to obtain a clear grayscale focused image , clear grayscale focused image , clear grayscale focused image , and clear grayscale focused image ; An eigenvalue matching unit, configured to perform eigenvalue matching on the images at the corresponding positions according to the regions and , and obtain the offset coordinates of the clear grayscale focused image , clear grayscale focused image , clear grayscale focused image , and clear grayscale focused image ; A calculation unit, configured to obtain the offset coordinates of the next channel relative to the first channel based on the offset coordinates of the clear grayscale focused image , clear grayscale focused image , clear grayscale focused image , and clear grayscale focused image ; A repetition unit, configured to repeatedly use the focus acquisition unit, the eigenvalue matching unit, and the calculation unit until the offset coordinates of all channels relative to the first channel and the complete images of all channels are obtained; An output unit, configured to register and stack the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image.

[0027] The present application further provides an electronic device, which may also be: including a memory and a processor; the memory is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the foregoing automatic registration and spatial alignment method based on multi-channel fluorescence images.

[0028] The present application further provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the foregoing automatic registration and spatial alignment method based on multi-channel fluorescence images are implemented.

[0029] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0030] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be implemented. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0031] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing devices to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0032] These computer program instructions can also be loaded onto a computer or other programmable data processing devices, so that a series of operation steps are executed on the computer or other programmable devices to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable devices provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0033] The above is only the preferred embodiment of the present application. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present application, several improvements and modifications can be made, and these improvements and modifications should also be regarded as the protection scope of the present application.

Claims

1. An automatic registration and spatial alignment method based on multi-channel fluorescence images, characterized in that Including: S1. Obtain n foci for focusing, set the first channel as the focusing channel, and save the clear grayscale focusing images collected during focusing; S2, Compare the gray-scale focus images with clear contrast , , ..., and select two focus images with obvious peaks and horizontal focus positions and , record the focus positions and of the focus images and ; S3, divide and into block diagrams, and traverse to select the regions with the most obvious two peaks in the histogram and ; S4, then use the focusing channel at the focusing position and perform focusing and capture the focused images to obtain clear grayscale focused images , clear grayscale focused image , clear grayscale focused image and clear grayscale focused image ; S5, according to the region and , perform eigenvalue matching on the images at the corresponding positions to obtain clear grayscale focus maps , clear grayscale focus map , clear grayscale focus map and clear grayscale focus map of the offset coordinates; S6, based on the clear grayscale focus map and the clear grayscale focus map and the clear grayscale focus map as well as the clear grayscale focus map to obtain the offset coordinates of the next channel relative to the first channel; S7. Repeat S4 - S6 until the offset coordinates of all channels relative to the first channel and the complete images of all channels are obtained; S8. Register and superimpose the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image.

2. An automatic registration and spatial alignment method based on multi-channel fluorescence images according to claim 1, characterized in that Obtain focus points for focusing, set the first channel as the focusing channel, and save the clear grayscale focusing images collected during focusing, specifically including: obtaining the focus points selected on the operation interface of the fluorescence scanner software , ,..., ; Focus on the image with the first channel as the focus channel, and focus at the focal position in sequence; Save the clear grayscale images collected during focusing clear grayscale images and and so on .

3. An automatic registration and spatial alignment method based on multi-channel fluorescence images according to claim 1, characterized in that Comparison a clear grayscale focus image 、 、...、 The grayscale histograms of..., select two focus images with obvious peaks and the focus positions being horizontal and , record the focus positions of the focus images and , specifically including: and ​ Comparison gray-scale histograms of several focus maps , , ..., ; Select two of the focus images. The histograms of both show two distinct and separated peaks, with deep and clear valleys, and the focal positions corresponding to the two images have the same coordinates in the X-axis or Y-axis direction. Record the focus images and of the focal positions and .

4. A method for automatic registration and spatial alignment based on multi-channel fluorescence images according to claim 1, characterized in that, Divide and into block diagrams, and traverse to select the regions with the most obvious two peaks in the histogram and , specifically including: Divide into 3x3 block diagrams, traverse the block diagrams and generate the grayscale histograms of each block diagram, and select the regions with the two most obvious peaks in the histograms ; Divide into 3x3 block diagrams, traverse the block diagrams and generate the grayscale histograms of each block diagram, and select the regions with the two most obvious peaks in the histograms .

5. An automatic registration and spatial alignment method based on multi-channel fluorescence images according to claim 1, characterized in that Then use the focus channel at the focus position and perform focusing and collect focused images to obtain clear grayscale focused images clear grayscale focused image clear grayscale focused image and clear grayscale focused image , specifically including: Use the focusing channel for focusing, and move the stage to position, and move from position to position for focusing and save the captured clear grayscale focusing image ; Move from to for focusing and save the captured clear grayscale focusing image ; Move from to position for focusing and save the captured clear grayscale focusing image ; Move from to for focusing and save the captured clear grayscale focusing image ; Among them and are the images corresponding to the first channel at the position of the next channel image, and are the images corresponding to the first channel at the position of the next channel image.

6. The automatic registration and spatial alignment method based on multi-channel fluorescence images according to claim 1, wherein According to the region and , perform eigenvalue matching on the images at the corresponding positions to obtain clear grayscale focus maps , clear grayscale focus map , clear grayscale focus map and clear grayscale focus map of the offset coordinates, specifically including: Match the eigenvalues of the regions of the image and to obtain the offset coordinates ; ; Match the eigenvalues of the regions of the image and to obtain the offset coordinates ; ; Match the eigenvalues of the regions of the image and to obtain the offset coordinates ;​ Match the eigenvalues of the regions of the image and to obtain the offset coordinates .

7. An automatic registration and spatial alignment method based on multi-channel fluorescence images according to claim 1, characterized in that, Based on the clear grayscale focus map and the clear grayscale focus map and the clear grayscale focus map and the clear grayscale focus map The offset coordinates of the next channel relative to the first channel are obtained based on the offset coordinates, specifically including: Calculate the average value of the four offset coordinates and the standard deviation of each offset coordinate , , , , and the formula is as follows: ； After removing the coordinate with the largest deviation among the four offset coordinates, calculate the average value, and the formula is as follows: ； This average value is the offset of the overall image of the next channel relative to the overall image of the first channel in the x-direction and y-direction.

8. An automatic registration and spatial alignment system based on multi-channel fluorescence images, characterized in that, Including: An acquisition unit for acquiring focus points for focusing, setting a first channel as a focusing channel, and saving the clear grayscale focusing images acquired during focusing; A comparison unit for comparing gray-scale focus images 、 、...、 and selecting two focus images with obvious peaks and horizontal focus positions and recording the focus positions and of the focus images and ; Selection unit, used to and be equally divided into block diagrams, and traverse to select the regions with the most obvious two peaks of the histogram and ; A focusing acquisition unit, which is used to perform focusing and acquire a focused image at the focusing position by using a focusing channel and to obtain a clear grayscale focused image a clear grayscale focused image a clear grayscale focused image and a clear grayscale focused image ; An eigenvalue matching unit for performing eigenvalue matching on the images at corresponding positions according to the region and to obtain clear grayscale focus maps , clear grayscale focus map , clear grayscale focus map and clear grayscale focus map and the offset coordinates thereof; A calculation unit for obtaining the offset coordinates of the next channel relative to the first channel based on the clear grayscale focus map and the clear grayscale focus map and the clear grayscale focus map and the clear grayscale focus map ; A repeating unit for repeatedly using the focus acquisition unit, eigenvalue matching unit, and calculation unit until the offset coordinates of all channels relative to the first channel and the complete images of all channels are obtained; An output unit for registering and superimposing the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image.

9. An electronic device, characterized in that, Including a memory and a processor; The memory is used to store instructions; The processor is used to operate according to the instructions to execute the steps of the method according to any one of claims 1 - 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method according to any one of claims 1 - 7.

Images