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

Abstract

The present application discloses a method, system, electronic device, and storage medium for automatic registration and spatial alignment based on multi-channel fluorescence images, including: setting the first channel as the focus channel, obtaining n focal points for focusing and collecting a focus map; comparing the grayscale histograms of the n images, selecting two images with two obvious peaks and horizontal focus positions, and recording the focus positions of the images; dividing the selected images into equal blocks, traversing the area with the two most obvious peaks in the histogram; focusing and image acquisition of the next channel image to obtain a clear grayscale image; matching the features of the images at the corresponding positions according to the area to obtain the offset coordinates of the clear grayscale image; obtaining the offset coordinates of the next channel based on the offset coordinates; and finally obtaining the offset coordinates of all channels; and registering and superimposing the complete images of all channels according to the obtained offsets to form a multi-channel fluorescence image. This ensures the accuracy of single channel splicing and saves time and cost.

Description

Technical Field

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

[0002] Multiplex immunofluorescence (mIF) technology has become an important tool in tumor microenvironment research, immunology, pathological diagnosis, and other fields due to its advantages such as multiplex detection, high resolution, and spatial information retention. A multi-channel fluorescence digital slide scanner is required to convert multi-channel fluorescent slides into digital images. Fluorescent slide samples are fixed to the scanner's platform. The scanning platform controls the translation of the tray along the X, Y, and Z axes to achieve precise movement and positioning of the slides, allowing them to complete image acquisition through the optical system above the scanning platform. By illuminating the slides with light sources of different wavelengths, fluorescence signals from different channels can be stimulated, allowing the slides to be scanned sequentially by channel to obtain multi-channel fluorescence images.

[0003] However, during the scanning process, due to the multiple movements of the platform, its cumulative errors and the rapid movement of the slices may cause inaccurate matching or inaccurate splicing of different channels of the multi-channel fluorescence image. Different channels of fluorescent tissue correspond to specific molecules or structures. If the matching is not accurate, it will cause loss of image information, which in turn affects the accuracy of medical diagnosis and pathological research. In the existing technology, because 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 cannot be matched through conventional image algorithms. The current technology basically uses different channels to switch 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 this application is to provide a method, system, electronic device and storage medium for automatic registration and spatial alignment 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] This application adopts the following technical solutions to achieve the above-mentioned invention objectives:

[0006] This application provides an automatic registration and spatial alignment method based on multi-channel fluorescence images, including:

[0007] S1, get n focus points for focusing, set the first channel as the focus channel, and save the collected focus A clear grayscale focused image;

[0008] S2, comparison A clear grayscale focused image 、 ,..., Grayscale histogram, select two focused images with obvious peaks and horizontal focus positions and , record the focused image and Focus position and ;

[0009] S3, will and Divide the graph into blocks and select the areas with the two most obvious peaks in the histogram and ;

[0010] S4, then use the focus channel to focus at the focus position and Focus and collect the focused image to obtain a clear grayscale focused image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image ;

[0011] S5, by region and , perform feature value matching on the images at the corresponding positions to obtain a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of

[0012] S6, based on a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of the next channel relative to the first channel are obtained;

[0013] 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;

[0014] S8, registering and superimposing the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image.

[0015] Further, obtain Focus the image with the first channel as the focus channel, and save the image captured during focus. A clear grayscale focused image, including:

[0016] Get the selected fluorescence scanner software interface Focus 、 ,..., ;

[0017] Focus the image using the first channel as the focus channel, focusing at the focus position in sequence;

[0018] Save the focus A clear grayscale image 、 ,..., .

[0019] Furthermore, the comparison A clear grayscale focused image 、 ,..., Grayscale histogram, select two focus images with obvious peaks and horizontal focus positions and , record the focus image and Focus position and , specifically including:

[0020] contrast Focus images 、 ,..., Grayscale histogram of ;

[0021] Select two focus images, whose histograms show two distinct and separate peaks, with deep and clear valleys, and the focus positions corresponding to the two images are consistent in the X-axis or Y-axis direction, and record the focus images. and Focus position and .

[0022] Further, and Divide the graph into blocks and select the areas with the two most obvious peaks in the histogram and , specifically including:

[0023] Will Divide the image into 3x3 blocks, traverse the blocks and generate the grayscale histogram of each block, and select the area with the two most obvious peaks in the histogram ;

[0024] Will Divide the image into 3x3 blocks, traverse the blocks and generate the grayscale histogram of each block, and select the area with the two most obvious peaks in the histogram .

[0025] Furthermore, use the focus channel to focus on the position and Focus and collect the focused image to obtain a clear grayscale focused image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image , specifically including:

[0026] Use the focus channel to focus and move the platform to Location, from Move to Focus the position and save the collected clear grayscale focus image ;

[0027] from Move to Focus the position and save the collected clear grayscale focus image ;

[0028] from Move to Focus the position and save the collected clear grayscale focus image ;

[0029] from Move to Focus the position and save the collected clear grayscale focus image ;

[0030] in and The next channel corresponds to the first channel Position image, and The next channel corresponds to the first channel Image of the location.

[0031] Furthermore, according to the region and , perform feature value matching on the images at the corresponding positions to obtain a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates include:

[0032] The image and Area Perform eigenvalue matching to obtain offset coordinates ;

[0033] The image and Area Perform eigenvalue matching to obtain offset coordinates ;

[0034] The image and Area Perform eigenvalue matching to obtain offset coordinates ;

[0035] The image and Area Perform eigenvalue matching to obtain offset coordinates .

[0036] Furthermore, based on the clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of the next channel relative to the first channel are obtained, specifically including:

[0037] Calculate the average of the four offset coordinates and the standard deviation of each offset coordinate 、 、 、 , the formula is as follows:

[0038] ;

[0039] The coordinate with the largest deviation among the four offset coordinates is removed and the average value is calculated. The formula is as follows:

[0040] ;

[0041] The average It is the offset of the next channel's overall image relative to the first channel's overall image in the x and y directions.

[0042] This application provides an automatic registration and spatial alignment system based on multi-channel fluorescence images, including:

[0043] Acquisition unit, used to obtain Focus the image with the first channel as the focus channel, and save the image captured during focus. A clear grayscale focused image;

[0044] Comparison unit, used for comparison A clear grayscale focused image 、 ,..., Grayscale histogram, select two focused images with obvious peaks and horizontal focus positions and , record the focused image and Focus position and ;

[0045] Select the unit to and Divide the graph into blocks and select the areas with the two most obvious peaks in the histogram and ;

[0046] Focus acquisition unit, used to reuse the focus channel at the focus position and Focus and collect the focused image to obtain a clear grayscale focused image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image ;

[0047] Eigenvalue matching unit, used to match the region and , perform feature value matching on the images at the corresponding positions to obtain a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of

[0048] Calculation unit for focusing based on clear grayscale image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of the next channel relative to the first channel are obtained;

[0049] a repeating unit, configured to repeatedly utilize the focus acquisition unit, the eigenvalue matching unit, and the calculation unit until offset coordinates of all channels relative to the first channel and complete images of all channels are obtained;

[0050] The output unit is used to register and superimpose the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image.

[0051] The present application provides an electronic device, including a memory and a processor;

[0052] The memory is used to store instructions;

[0053] The processor is configured to operate according to the instructions to execute the steps of the aforementioned automatic registration and spatial alignment method based on multi-channel fluorescence images.

[0054] The present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the aforementioned automatic registration and spatial alignment method based on multi-channel fluorescence images.

[0055] The beneficial effects of this application are as follows:

[0056] This application scans images channel by channel. After scanning all the small images in each channel, they are stitched together into a large image. The next channel is scanned, and finally the images from different channels are registered and superimposed. This method simplifies the stitching process, ensuring the accuracy of individual channel stitching, while eliminating errors between different channels through the registration algorithm. Switching between different channels only needs to be done once, which greatly saves time and cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flowchart of an automatic registration and spatial alignment method based on multi-channel fluorescence images according to an embodiment of the present application;

[0058] Figure 2 A schematic diagram of a fluorescence multi-channel method for automatic registration and spatial alignment of multi-channel fluorescence images provided in accordance with an embodiment of the present application;

[0059] Figure 3 Schematic diagram of image overlap of different channels in an ideal state and an 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;

[0060] Figure 4 This is an image grayscale distribution diagram that meets the selection requirements of an automatic registration and spatial alignment method based on multi-channel fluorescence images provided in an embodiment of the present application. DETAILED DESCRIPTION

[0061] During the fluorescence image scanning process, the same location on the same slice is excited and scanned multiple times using light of different wavelengths. Each wavelength represents a channel, and the overall image of each channel is composed of multiple small images stitched together. The scanning platform, optical system, and camera acquisition system work together to complete image acquisition. During the image acquisition process, the movement of the platform inevitably causes offset. When the offset error accumulates to a certain level, it will affect the quality of the generated fluorescence image. In the prior art, at the same location, small images are scanned sequentially using different channels to reduce the offset between channels. This can eliminate the error between small images in different channels at the same location. In software, this method accumulates the error into the overall image of the same channel, which places high demands on the stitching algorithm and makes stitching difficult. In hardware, the frequent switching between different channels increases the scanning time cost, which is not conducive to the realization of rapid scanning of fluorescence slices. Based on this, the present invention scans the image sequentially by channel. After all the small images in each channel are scanned, they are stitched into a large image. Then the next channel is scanned. Finally, the images of different channels are aligned and superimposed. In terms of software, this method not only simplifies the difficulty of stitching and ensures the accuracy of stitching of a single channel, but also eliminates the errors between different channels through the registration algorithm; in terms of hardware, different channels only need to be switched once, greatly saving time and cost.

[0062] like Figures 1 to 4 As shown, the present application provides an automatic registration and spatial alignment method based on multi-channel fluorescence images, which specifically includes the following steps:

[0063] S1: Acquisition Focus the image with the first channel as the focus channel, and save the image captured during focus. A clear grayscale focused image;

[0064] The S1 is specifically:

[0065] Set the first channel as the focus channel to ensure that the focus images collected in subsequent multiple focus acquisitions have the same source. Select focus 、 ,..., , focus at the focus position in turn, and save the data collected during focus A clear grayscale focus image 、 ,..., . Then the image acquisition of this channel is completed.

[0066] S2, comparison A clear grayscale focused image 、 ,..., Grayscale histogram, select two focus images with obvious peaks and horizontal focus positions and , record the focus image and Focus position and ;

[0067] Said S2 is specifically:

[0068] contrast images 、 ,..., Select two images from the grayscale histogram, and their histograms show two obvious and separated peaks, with deep and clear troughs, and the focus positions corresponding to the two images are consistent in the X-axis or Y-axis direction, and record the focus map. and Focus position and The histogram shows two distinct and separate peaks, representing the foreground and background regions. The foreground and background are easily separated, allowing for better feature point calculation. The consistency of the focal positions corresponding to the two images in the X-axis or Y-axis coordinates allows the mobile platform to move in only one direction when focusing, reducing movement errors in the other direction.

[0069] S3: and Divide the graph into blocks and select the areas with the two most obvious peaks in the histogram and ;

[0070] The S3 is specifically:

[0071] Will Divide the image into 3x3 blocks, traverse the blocks and generate the grayscale histogram of each block, and select the area with the two most obvious peaks in the histogram ;Will Divide the image into 3x3 blocks, traverse the blocks and generate the grayscale histogram of each block, and select the area with the two most obvious peaks in the histogram ,Compared to the entire image, the block image information is simpler, ,easy to find feature points, and the registration algorithm has better ,effects.

[0072] S4: Use the focus channel again to focus at the focus position and Focus and collect the focused image to obtain a clear grayscale focused image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image ;

[0073] The S4 is specifically:

[0074] Focus again, use the focus channel to focus and collect the focus image, move the platform to Location, from Move to Focus the position and save the collected clear grayscale focus image ,from Move to Focus the position and save the collected clear grayscale focus image ,from Move to Focus the position and save the collected clear grayscale focus image ,from Move to Focus the position and save the collected clear grayscale focus image .in and The next channel corresponds to the first channel Position image, and The next channel corresponds to the first channel The purpose of moving from one focus position to another and then focusing is to simulate the movement of the platform during normal focusing and control other environmental factors when collecting images at the same focus each time.

[0075] S5: According to the region and , perform feature value matching on the images at the corresponding positions to obtain a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of

[0076] The S5 is specifically:

[0077] Focus image and Area Perform eigenvalue matching to obtain offset coordinates , focus image and Area Perform eigenvalue matching to obtain offset coordinates , focus image and Area Perform eigenvalue matching to obtain offset coordinates , focus image and Area Perform eigenvalue matching to obtain offset coordinates Specifically, the Scale Invariant Feature Transform (SIFT) algorithm is used. The core idea of the 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 keypoints 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. The algorithm then assigns a principal direction to each keypoint and generates a rotationally invariant 128-dimensional feature description vector based on the gradient direction statistics within the keypoint's neighborhood. This vector is constructed by partitioning a 16×16 neighborhood into 4×4 subregions and calculating the gradient histograms of eight directions in each subregion. During the feature matching stage, a nearest neighbor search combined with a ratio test is used to find reliable matching point pairs by comparing the Euclidean distance between feature vectors. A random sampling consensus algorithm is then used to eliminate false matches to further improve matching accuracy. Finally, through coordinate transformation analysis of correctly matched point pairs, the offset coordinates between images can be accurately calculated.

[0078] S6: Based on clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of the next channel relative to the first channel are obtained;

[0079] The S6 is specifically:

[0080] Calculate the average of the four offset coordinates and the standard deviation of each offset coordinate 、 、 、 ,in:

[0081] ;

[0082] In order to reduce the error and ensure the accuracy of the results, the coordinate with the largest deviation among the four offset coordinates is removed and then the average value is calculated. The offset coordinate with the highest standard deviation is the outlier that deviates more from the mean. The average value of the remaining three offset coordinates is obtained. ,in:

[0083] ;

[0084] This average It is the offset of the next channel overall image relative to the first channel overall image in the x and y directions. Specifically, calculating the dynamic standard deviation and eliminating outliers can effectively improve the robustness of the eigenvalue matching algorithm.

[0085] 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;

[0086] The S7 is specifically:

[0087] If the image of the next channel has not been collected yet, steps S4-S6 are repeated until the offset calculation of all channels is completed.

[0088] S8, registering and superimposing the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image;

[0089] The S8 is specifically:

[0090] A unified image spatial reference system is established using the first channel (usually DAPI nuclear stain or the clearest channel) as the reference coordinate system. Specific pseudo-color mappings are preset for other channels before image acquisition. These colors strictly correspond to the emission wavelength characteristics of the specific fluorescent dyes used during imaging: for example, DAPI nuclear stain typically excites at 340 nm and emits at 488 nm, resulting in a blue image; FITC excites at 495 nm and emits at 518 nm, resulting in a green image; Cy3 excites at 550 nm and emits at 570 nm, resulting in an orange-red image; and Cy5 excites at 650 nm and emits at 670 nm, resulting in a deep red image. After spatial registration of each channel, the images are combined according to the preset wavelength-color mapping, ultimately generating a multi-channel fluorescence composite image with well-defined spectral characteristics.

[0091] The present application also provides an automatic registration and spatial alignment system based on multi-channel fluorescence images, comprising:

[0092] Acquisition unit, used to obtain Focus the image with the first channel as the focus channel, and save the image captured during focus. A clear grayscale focused image;

[0093] Comparison unit, used for comparison A clear grayscale focused image 、 ,..., Grayscale histogram, select two focused images with obvious peaks and horizontal focus positions and , record the focused image and Focus position and ;

[0094] Select the unit to and Divide the graph into blocks and select the areas with the two most obvious peaks in the histogram and ;

[0095] Focus acquisition unit, used to reuse the focus channel at the focus position and Focus and collect the focused image to obtain a clear grayscale focused image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image ;

[0096] Eigenvalue matching unit, used to match the region and , perform feature value matching on the images at the corresponding positions to obtain a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of

[0097] Calculation unit for focusing based on clear grayscale image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of the next channel relative to the first channel are obtained;

[0098] a repeating unit, configured to repeatedly utilize the focus acquisition unit, the eigenvalue matching unit, and the calculation unit until offset coordinates of all channels relative to the first channel and complete images of all channels are obtained;

[0099] The output unit is used to register and superimpose the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image.

[0100] The present application also provides an electronic device, which may include: a memory and a processor; the memory is used to store instructions;

[0101] The processor is configured to operate according to the instructions to execute the steps of the aforementioned automatic registration and spatial alignment method based on multi-channel fluorescence images.

[0102] The present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the aforementioned method for automatic registration and spatial alignment based on multi-channel fluorescence images.

[0103] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0104] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0105] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0106] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0107] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for automatic registration and spatial alignment of multi-channel fluorescence images, characterized in that: include: S1, get n focus points for focusing, set the first channel as the focus channel, and save the collected focus A clear grayscale focused image; S2, comparison A clear grayscale focused image 、 ,..., Grayscale histogram, select two focused images with obvious peaks and horizontal focus positions and , record the focused image and Focus position and ; S3, will and Divide the graph into blocks and select the areas with the two most obvious peaks in the histogram and ; S4, then use the focus channel to focus at the focus position and Focus and collect the focused image to obtain a clear grayscale focused image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image ; S5, by region and , perform feature value matching on the images at the corresponding positions to obtain a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of S6, based on a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of the next channel relative to the first channel are obtained; 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, registering and superimposing the images of all channels according to the obtained offset coordinates to form a multi-channel fluorescence image.

2. The method for automatic registration and spatial alignment based on multi-channel fluorescence images according to claim 1, characterized in that: Get Focus the image with the first channel as the focus channel, and save the image captured during focus. A clear grayscale focused image is obtained by: obtaining the selected Focus 、 ,..., ; Focus the image using the first channel as the focus channel, focusing at the focus position in sequence; Save the focus A clear grayscale image 、 ,..., .

3. The method for automatic registration and spatial alignment based on multi-channel fluorescence images according to claim 1, characterized in that: contrast A clear grayscale focused image 、 ,..., Grayscale histogram, select two focus images with obvious peaks and horizontal focus positions and , record the focused image and Focus position and , specifically including: contrast Focus images 、 ,..., Grayscale histogram of ; Select two focus images, whose histograms show two distinct and separate peaks, with deep and clear valleys, and the focus positions corresponding to the two images are consistent in the X-axis or Y-axis direction, and record the focus images. and Focus position and .

4. The method for automatic registration and spatial alignment based on multi-channel fluorescence images according to claim 1, characterized in that: Will and Divide the graph into blocks and select the areas with the two most obvious peaks in the histogram and , specifically including: Will Divide the image into 3x3 blocks, traverse the blocks and generate the grayscale histogram of each block, and select the area with the two most obvious peaks in the histogram ; Will Divide the image into 3x3 blocks, traverse the blocks and generate the grayscale histogram of each block, and select the area with the two most obvious peaks in the histogram .

5. The method for automatic registration and spatial alignment based on multi-channel fluorescence images according to claim 1, characterized in that: Then use the focus channel to focus on the position and Focus and collect the focused image to obtain a clear grayscale focused image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image , specifically including: Use the focus channel to focus and move the platform to Location, from Move to Focus the position and save the collected clear grayscale focus image ; from Move to Focus the position and save the collected clear grayscale focus image ; from Move to Focus the position and save the collected clear grayscale focus image ; from Move to Focus the position and save the collected clear grayscale focus image ; in and The next channel corresponds to the first channel Position image, and The next channel corresponds to the first channel Image of the location.

6. The method for automatic registration and spatial alignment based on multi-channel fluorescence images according to claim 1, characterized in that: According to the region and , perform feature value matching on the images at the corresponding positions to obtain a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates include: The image and Area Perform eigenvalue matching to obtain offset coordinates ; The image and Area Perform eigenvalue matching to obtain offset coordinates ; The image and Area Perform eigenvalue matching to obtain offset coordinates ; The image and Area Perform eigenvalue matching to obtain offset coordinates .

7. The method for automatic registration and spatial alignment based on multi-channel fluorescence images according to claim 1, characterized in that: Based on clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of the next channel relative to the first channel are obtained, specifically including: Calculate the average of the four offset coordinates and the standard deviation of each offset coordinate 、 、 、 , the formula is as follows: ； The coordinate with the largest deviation among the four offset coordinates is removed and the average value is calculated. The formula is as follows: ； The average It is the offset of the next channel's overall image relative to the first channel's overall image in the x and y directions.

8. An automatic registration and spatial alignment system based on multi-channel fluorescence images, characterized in that: include: Acquisition unit, used to obtain Focus the image with the first channel as the focus channel, and save the image captured during focus. A clear grayscale focused image; Comparison unit, used for comparison A clear grayscale focused image 、 ,..., Grayscale histogram, select two focused images with obvious peaks and horizontal focus positions and , record the focused image and Focus position and ; Select the unit to and Divide the graph into blocks and select the areas with the two most obvious peaks in the histogram and ; Focus acquisition unit, used to reuse the focus channel at the focus position and Focus and collect the focused image to obtain a clear grayscale focused image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image ; Eigenvalue matching unit, used to match the region and , perform feature value matching on the images at the corresponding positions to obtain a clear grayscale focus image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of Calculation unit for focusing based on clear grayscale image , clear grayscale focus image , clear grayscale focus image And clear grayscale focus image The offset coordinates of the next channel relative to the first channel are obtained; a repeating unit, configured to repeatedly utilize the focus acquisition unit, the eigenvalue matching unit, and the calculation unit until offset coordinates of all channels relative to the first channel and complete images of all channels are obtained; The output unit is used to register and superimpose 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 memory and processor; The memory is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1 to 7.

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

Images