Dark-field microdroplet image recognition method and device, computer device and storage medium

Abstract

The application relates to a dark-field microdroplet image recognition method and device of a nucleic acid molecule, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring a dark-field microdroplet image, extracting foreground markers, background markers and gradient amplitudes of the image from the dark-field microdroplet image; superimposing the foreground markers, the background markers and the gradient amplitudes of the image to obtain a gradient amplitude image; and segmenting the gradient amplitude image to obtain images of each microdroplet in the dark-field microdroplet image. The method can improve the accuracy of dark-field microdroplet recognition.

Description

Technical Field

[0001] This application relates to the field of droplet digital PCR technology, and in particular to a method, apparatus, computer equipment, and storage medium for dark-field droplet image recognition. Background Technology

[0002] Droplet digital polymerase chain reaction (DPCR) is a third-generation digital PCR technology, a method for absolute quantification of nucleic acid molecules, allowing direct measurement of the copy number of a target gene in a sample. This technology uses a digital PCR chip to dilute nucleic acid samples and randomly disperse them into tens of thousands of water-in-oil droplets. The DNA sample in each droplet is then bound to a specific fluorescent label for PCR. After the PCR reaction, the droplets are photographed using a CCD camera under bright-field and dark-field conditions, respectively, yielding bright-field and dark-field droplet images. Finally, image recognition and detection are performed on both types of images.

[0003] Obtaining dark-field fluorescence droplet images requires a completely dark environment. First, the droplets are illuminated with an excitation light source to make positive droplets luminescent. Due to the limited intensity of the excitation light source and the non-uniformity of illumination, the brightness and contrast of dark-field fluorescence droplet images are low. Furthermore, because the droplets are densely packed, there is adhesion between them, leading to low computational accuracy in current droplet recognition algorithms. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer device, and storage medium for dark-field droplet image recognition that can improve the accuracy of droplet identification, in order to address the aforementioned technical problems.

[0005] In a first aspect, this application provides a method for dark-field droplet image recognition of nucleic acid molecules. The method includes:

[0006] Acquire the dark field droplet image, and extract the foreground markers, background markers, and gradient magnitude of the image from the dark field droplet image;

[0007] The foreground marker, background marker, and gradient magnitude of the image are superimposed to obtain a gradient magnitude image;

[0008] The gradient magnitude image is segmented to obtain the image of each droplet in the dark field droplet image.

[0009] In one embodiment, extracting foreground markers, background markers, and gradient magnitudes from the dark-field droplet image includes:

[0010] Calculate the segmentation thresholds for foreground and background markers in the dark field droplet image based on the dark field droplet image;

[0011] The dark field droplet image is segmented according to the segmentation threshold to obtain a binary image of the foreground in the dark field droplet image;

[0012] The binary image of the foreground is eroded, and then the erosion result is dilated to obtain the processed binary image of the foreground.

[0013] Identify the maxima region in the processed foreground binary image and calculate the area of ​​the maxima region;

[0014] The regions in the processed foreground binary image whose area differs from the area of ​​the maximum region by a preset difference are removed to obtain the foreground marker.

[0015] In one embodiment, after eroding the binary image of the foreground and then dilating the erosion result to obtain the processed binary image of the foreground, the process includes:

[0016] For each first pixel in the foreground binary image, the distance between it and the second pixel that is closest to the first pixel is used to obtain a pixel image represented by distance;

[0017] The watershed algorithm is used to determine the boundary lines in the pixel image;

[0018] The background marker is obtained from the dark field droplet image based on the boundary line.

[0019] In one embodiment, the step of extracting foreground markers, background markers, and gradient magnitudes from the dark-field droplet image further includes:

[0020] The dark field droplet image was processed using an adaptive H-minima transform;

[0021] The gradient magnitude of the dark field droplet image is obtained by calculating the gradient magnitude of the dark field droplet image using the Sobel operator.

[0022] In one embodiment, prior to acquiring the dark-field droplet image, the process includes:

[0023] Obtain the original dark-field droplet image and perform Gaussian filtering on the original dark-field droplet image.

[0024] In one embodiment, obtaining the original dark-field droplet image and then performing Gaussian filtering on the original dark-field droplet image includes:

[0025] The original dark field droplet image after Gaussian filtering is enhanced to obtain the dark field droplet image.

[0026] Secondly, this application also provides a dark-field droplet image recognition device for nucleic acid molecules. The device includes:

[0027] An extraction module is used to acquire the dark field droplet image and extract foreground markers, background markers, and gradient magnitudes from the dark field droplet image.

[0028] The overlay module is used to overlay the foreground marker, background marker, and gradient magnitude of the image to obtain a gradient magnitude image;

[0029] The segmentation module is used to segment the gradient magnitude image to obtain images of each droplet in the dark field droplet image.

[0030] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0031] Acquire the dark field droplet image, and extract the foreground markers, background markers, and gradient magnitude of the image from the dark field droplet image;

[0032] The foreground marker, background marker, and gradient magnitude of the image are superimposed to obtain a gradient magnitude image;

[0033] The gradient magnitude image is segmented to obtain the image of each droplet in the dark field droplet image.

[0034] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0035] Acquire the dark field droplet image, and extract the foreground markers, background markers, and gradient magnitude of the image from the dark field droplet image;

[0036] The foreground marker, background marker, and gradient magnitude of the image are superimposed to obtain a gradient magnitude image;

[0037] The gradient magnitude image is segmented to obtain the image of each droplet in the dark field droplet image.

[0038] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0039] Acquire the dark field droplet image, and extract the foreground markers, background markers, and gradient magnitude of the image from the dark field droplet image;

[0040] The foreground marker, background marker, and gradient magnitude of the image are superimposed to obtain a gradient magnitude image;

[0041] The gradient magnitude image is segmented to obtain the image of each droplet in the dark field droplet image.

[0042] The aforementioned method, apparatus, computer equipment, storage medium, and computer program products for dark-field droplet image recognition of nucleic acid molecules first acquire the dark-field droplet image, then extract foreground markers, background markers, and the gradient amplitude of the image from the dark-field droplet image, eliminating the influence of background noise during image processing. Next, the foreground markers, background markers, and the gradient amplitude of the image are superimposed to obtain a gradient amplitude image, resulting in an image more conducive to segmentation. Finally, segmentation is performed based on the gradient amplitude image to obtain images of each droplet in the dark-field droplet image. This significantly improves the segmentation accuracy, thereby enhancing the accuracy of droplet recognition in dark-field droplet images. Attached Figure Description

[0043] Figure 1 This is an application environment diagram of a dark-field droplet image recognition method for nucleic acid molecules in one embodiment;

[0044] Figure 2 This is a flowchart illustrating a method for dark-field droplet image recognition of nucleic acid molecules in one embodiment;

[0045] Figure 3 This is a flowchart illustrating a method for dark-field droplet image recognition of nucleic acid molecules in another embodiment;

[0046] Figure 4 This is a structural block diagram of a dark-field droplet image recognition device for nucleic acid molecules in one embodiment;

[0047] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0049] The dark-field droplet image recognition method for nucleic acid molecules provided in this application can be applied to, for example... Figure 1In the illustrated application environment, the system includes a camera device 102 and a server 104. The camera device 102 can communicate with the server 104 via wired or wireless means. A data storage system can store the image data that the server 104 needs to process. The data storage system can be integrated onto the server 104, or it can be placed in the cloud or on another network server. The camera device 102 and the server 104 can also be integrated into a single terminal. The camera device 102 is used to photograph dark-field droplets of nucleic acid molecules (in a dark environment, the droplets are illuminated with an excitation light source to make positive droplets glow, and then photographed by the camera device 102), obtaining a droplet image of each dark-field droplet of the nucleic acid molecule. The server 104 identifies the droplets based on the received droplet images to obtain a droplet image. The server 104 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, or other devices or platforms with image processing capabilities. In practice, after the camera device 102 obtains the microdroplet image, the user can also copy or cut the dark field microdroplet image captured by the camera device 102 to the server 104 via a USB flash drive or card reader.

[0050] In one embodiment, such as Figure 2 As shown, a method for dark-field droplet image recognition of nucleic acid molecules is provided, including the following steps:

[0051] Step S210: Obtain the dark field droplet image, and extract the foreground marker, background marker, and gradient magnitude of the image from the dark field droplet image.

[0052] Droplet digital polymerase chain reaction (DPCR) is a third-generation digital PCR technology, a method for absolute quantification of nucleic acid molecules, allowing direct measurement of the copy number of a target gene in a sample. This technology uses a digital PCR chip to dilute nucleic acid samples and randomly disperse them into tens of thousands of water-in-oil droplets. The DNA sample in each droplet is then bound to a specific fluorescent label for PCR. After the PCR reaction, the droplets are photographed using a microscope and a CCD camera (or other cameras) in both dark-field and dark-field environments, resulting in dark-field and dark-field fluorescent droplet images. Obtaining dark-field fluorescent droplet images requires a completely dark environment. First, the droplets are illuminated with an excitation light source to cause positive droplets to glow, resulting in a long exposure time. Second, due to the limited intensity of the excitation light source and the uneven illumination, the brightness and contrast of the dark-field fluorescent droplet images are low. Finally, the dense arrangement of the droplets leads to adhesion between them, causing inaccurate droplet identification in existing algorithms.

[0053] In this embodiment, the method is applied to Figure 1Taking the server as an example, the server can control the camera device to take pictures of the droplets and obtain dark-field droplet images. After obtaining the dark-field droplet images taken by the camera device, the server extracts the foreground markers, background markers and the gradient magnitude of the image from the dark-field droplet images.

[0054] As an example, extracting foreground markers, background markers, and gradient magnitudes from the dark-field droplet image includes:

[0055] Calculate the segmentation thresholds for foreground and background markers in the dark field droplet image based on the dark field droplet image;

[0056] The dark field droplet image is segmented according to the segmentation threshold to obtain a binary image of the foreground in the dark field droplet image;

[0057] The binary image of the foreground is eroded, and then the erosion result is dilated to obtain the processed binary image of the foreground.

[0058] Identify the maxima region in the processed foreground binary image and calculate the area of ​​the maxima region;

[0059] The regions in the processed foreground binary image whose area differs from the area of ​​the maximum region by a preset difference are removed to obtain the foreground marker.

[0060] Specifically, the process of extracting foreground markers from the dark-field droplet image may include: calculating the segmentation thresholds for foreground and background markers in the dark-field droplet image. Specifically, this may involve: converting the dark-field droplet image into a grayscale image, and then using each grayscale value in the grayscale image as a threshold to segment all grayscale values ​​into two parts. For example, assuming the grayscale value is 0-m, first use the grayscale value 0 to segment 0-m, obtaining two grayscale values: (0) and (1-m); then use the grayscale value 1 to segment 0-m, obtaining two grayscale values: (0, 1) and (2-m); and so on. Then, for each grayscale value, calculate the proportions p1 and p2 of the two grayscale values ​​corresponding to each part, the average grayscale values ​​V1 and V2 of each part, and the overall average grayscale value V. For example, for grayscale 0, p1 is the proportion of 0 to 0-m; p2 is the proportion of 1 to 0-m; V1 is the average grayscale value of (0); V2 is the average grayscale value of (1-m); and V is the average grayscale value of 0-m. Calculate the variance of each grayscale value based on p1, p2, V1, V2, and V. The grayscale value with the largest variance is the segmentation threshold. The segmentation threshold can also be calculated using other methods in existing technologies, such as Otsu threshold segmentation, which will not be elaborated here.

[0061] After obtaining the segmentation threshold, the dark field droplet image is segmented using the segmentation threshold to obtain a binary image of the foreground in the dark field droplet image, that is, a binary image that separates the droplets from the background.

[0062] Then, the binary image of the foreground is eroded. For ease of understanding, let's assume the binary image of the foreground is image 1. Image 1 contains multiple other tiny parts adjacent to it. Eroding image 1 can be understood as slimming it down, and then dilating the slimmed-down image 1 back to its original size, thus eliminating the parts that affect image 1 and obtaining a clear image 1. After eroding and dilating the binary image of the foreground, a clear binary image of the foreground (i.e., the processed binary image of the foreground) is obtained. The specific process of eroding and dilating the image can be achieved using other methods, such as using morphological opening operations to calculate the process on dark-field droplet images, which will not be elaborated here.

[0063] After calculating the segmentation threshold, using the segmentation threshold to segment the dark field droplet image, and performing erosion and dilation processing on the binary image, these processes can eliminate background noise and initially segment the adhering droplets.

[0064] Then, in the processed binary image of the foreground, the maximum region is determined, and the area of ​​the maximum region and the area of ​​other regions are calculated respectively. Regions whose area is greater than the area of ​​the maximum region are removed (i.e., regions whose area is much smaller than the area of ​​the maximum region are removed). In this way, the content that affects the recognition accuracy of the image is further removed, and the final image is used as the foreground marker.

[0065] As another embodiment, extracting foreground markers, background markers, and gradient magnitudes from the dark-field droplet image includes:

[0066] Calculate the segmentation thresholds for foreground and background markers in the dark field droplet image based on the dark field droplet image;

[0067] The dark field droplet image is segmented according to the segmentation threshold to obtain a binary image of the foreground in the dark field droplet image;

[0068] The binary image of the foreground is eroded, and then the erosion result is dilated to obtain the processed binary image of the foreground.

[0069] For each first pixel in the foreground binary image, the distance between it and the second pixel that is closest to the first pixel is used to obtain a pixel image represented by distance;

[0070] The watershed algorithm is used to determine the boundary lines in the pixel image;

[0071] The background marker is obtained from the dark field droplet image based on the boundary line.

[0072] Specifically, the processes of calculating the segmentation threshold, segmenting the dark-field droplet image using the segmentation threshold, and performing erosion and dilation processing on the binary image during the extraction of background markers from the dark-field droplet image are the same as in the previous embodiment. After obtaining the processed foreground binary image, a distance transformation is first performed on the foreground binary image. That is, in two-dimensional space, points with a pixel value of 1 in the foreground binary image are called the first pixel, and points with a pixel value of 0 are called the second pixel. The distance transformation calculates the distance between each first pixel and its nearest second pixel in the image. The result is a grayscale image similar to the droplet image, but the grayscale value only appears in the foreground first pixel, and the grayscale value of pixels farther away from the background edge is larger. Then, the watershed algorithm is used to obtain the boundary line between adjacent regions, and the background marker of the dark-field droplet image is obtained based on the obtained boundary line.

[0073] In yet another embodiment, extracting foreground markers, background markers, and gradient magnitudes from the dark-field droplet image includes:

[0074] The dark field droplet image was processed using an adaptive H-minima transform;

[0075] The gradient magnitude of the dark field droplet image is obtained by calculating the gradient magnitude of the dark field droplet image using the Sobel operator.

[0076] Specifically, the process of extracting the gradient magnitude of the image from the dark-field droplet image may include: applying an adaptive H-minima transform to the dark-field droplet image, where the adaptive H-minima transform is a segmentation algorithm; and then using the Sobel operator to calculate the gradient magnitude of the dark-field droplet image after processing with the adaptive H-minima transform. The Sobel operator is a discrete differentiation operator. It combines Gaussian smoothing and differentiation to calculate the approximate gradient of the image's grayscale function.

[0077] Step S220: The foreground marker, background marker, and gradient magnitude of the image are superimposed to obtain a gradient magnitude image.

[0078] After obtaining the foreground markers, background markers, and the gradient magnitude of the image, these values ​​are superimposed to obtain a gradient magnitude image. This results in a gradient magnitude image with clear layers and eliminated background noise.

[0079] Step S230: Segment the gradient amplitude image to obtain images of each droplet in the dark field droplet image.

[0080] Based on the gradient magnitude image obtained above, a segmentation algorithm, such as the watershed algorithm, is used to segment the image, thereby obtaining the image of each droplet in the dark field droplet image.

[0081] The aforementioned method, apparatus, computer equipment, storage medium, and computer program products for dark-field droplet image recognition of nucleic acid molecules first acquire the dark-field droplet image, then extract foreground markers, background markers, and the gradient amplitude of the image from the dark-field droplet image, eliminating the influence of background noise during image processing. Next, the foreground markers, background markers, and the gradient amplitude of the image are superimposed to obtain a gradient amplitude image, resulting in an image more conducive to segmentation. Finally, segmentation is performed based on the gradient amplitude image to obtain images of each droplet in the dark-field droplet image. This significantly improves the segmentation accuracy, thereby enhancing the accuracy of droplet recognition in dark-field droplet images.

[0082] In one embodiment, based on the above embodiments, the method further includes:

[0083] Step S240: Obtain the original dark field droplet image and perform Gaussian filtering on the original dark field droplet image.

[0084] Step S250: Brightness enhancement is performed on the original dark field droplet image after Gaussian filtering to obtain the dark field droplet image.

[0085] Gaussian filtering is used to denoise the image to prevent noise from affecting subsequent image processing. Then, adaptive histogram equalization is used to enhance the image, improve the contrast and brightness, especially the brightness of fluorescent droplets in dark corners (adaptive histogram equalization algorithm increases contrast and brightness), and improve the accuracy of subsequent analysis.

[0086] In one embodiment, see Figure 3 The dark-field droplet image recognition method for this nucleic acid molecule may include:

[0087] The original dark-field droplet image is obtained, followed by image preprocessing to obtain a dark-field droplet image. Then, foreground markers, background markers, and gradient magnitudes are extracted from the dark-field droplet image. The extracted foreground markers, background markers, and gradient magnitudes are superimposed to obtain a gradient magnitude image. Finally, the gradient magnitude image is segmented using the watershed algorithm, and the droplets in the image are counted.

[0088] The foreground marker extraction process includes: calculating a segmentation threshold for the image using Otsu thresholding; using this threshold to segment the dark-field droplet image to obtain a binary image of the foreground. Then, morphological opening operations are applied to the dark-field droplet image to eliminate background noise and preliminarily segment adhered droplets. Next, maxima regions are identified in the processed binary foreground image. The areas of these maxima regions and other regions are calculated. Regions whose area differs from the area of ​​the maxima region by a preset difference are removed (i.e., regions with areas much smaller than the maxima region are removed). This further eliminates elements that affect recognition accuracy, and the final image serves as the foreground marker.

[0089] Background marker extraction includes: calculating a segmentation threshold for the image using Otsu thresholding; using this threshold to segment the dark-field droplet image to obtain a binary image of the foreground; then processing the dark-field droplet image using morphological opening to eliminate background noise and preliminarily segment adhering droplets; and finally performing a distance transform on the foreground binary image, where pixels with a value of 1 are designated as first pixels and pixels with a value of 0 as second pixels. The distance transform calculates the distance between each first pixel and its nearest second pixel, resulting in a grayscale image similar to the droplet image, but with grayscale values ​​only appearing in the foreground first pixels, and pixels farther from the background edge having larger grayscale values. Finally, the watershed algorithm is used to obtain the boundary lines between adjacent regions, and the background markers of the dark-field droplet image are obtained based on these boundary lines.

[0090] Gradient magnitude extraction includes: applying an adaptive H-minima transform to the dark field droplet image, and then using the Sobel operator to calculate the gradient magnitude of the dark field droplet image after the adaptive H-minima transform.

[0091] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0092] Based on the same inventive concept, this application also provides a device for dark-field droplet image recognition of nucleic acid molecules as described above. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the dark-field droplet image recognition device for nucleic acid molecules provided below can be found in the limitations of the dark-field droplet image recognition method for nucleic acid molecules described above, and will not be repeated here.

[0093] In one embodiment, such as Figure 4 As shown, a dark-field droplet image recognition device for nucleic acid molecules is provided, comprising:

[0094] Extraction module 410 is used to acquire the dark field droplet image and extract foreground markers, background markers, and gradient magnitude of the image from the dark field droplet image;

[0095] The overlay module 420 is used to overlay the foreground marker, the background marker, and the gradient magnitude of the image to obtain a gradient magnitude image;

[0096] The segmentation module 430 is used to segment the gradient amplitude image to obtain the image of each droplet in the dark field droplet image.

[0097] In one embodiment, the extraction module 410 is further configured to:

[0098] Calculate the segmentation thresholds for foreground and background markers in the dark field droplet image based on the dark field droplet image;

[0099] The dark field droplet image is segmented according to the segmentation threshold to obtain a binary image of the foreground in the dark field droplet image;

[0100] The binary image of the foreground is eroded, and then the erosion result is dilated to obtain the processed binary image of the foreground.

[0101] Identify the maxima region in the processed foreground binary image and calculate the area of ​​the maxima region;

[0102] The regions in the processed foreground binary image whose area differs from the area of ​​the maximum region by a preset difference are removed to obtain the foreground marker.

[0103] In one embodiment, the extraction module 410 is further configured to:

[0104] For each first pixel in the foreground binary image, the distance between it and the second pixel that is closest to the first pixel is used to obtain a pixel image represented by distance;

[0105] The watershed algorithm is used to determine the boundary lines in the pixel image;

[0106] The background marker is obtained from the dark field droplet image based on the boundary line.

[0107] In one embodiment, the extraction module 410 is further configured to:

[0108] The dark field droplet image was processed using an adaptive H-minima transform;

[0109] The gradient magnitude of the dark field droplet image is obtained by calculating the gradient magnitude of the dark field droplet image using the Sobel operator.

[0110] In one embodiment, the dark-field droplet image recognition device for nucleic acid molecules further includes:

[0111] A Gaussian filtering module (not shown) is used to obtain the original dark-field droplet image and perform Gaussian filtering on the original dark-field droplet image.

[0112] In one embodiment, the dark-field droplet image recognition device for nucleic acid molecules further includes:

[0113] A brightness enhancement processing module (not shown in the figure) is used to enhance the brightness of the original dark field droplet image after Gaussian filtering to obtain the dark field droplet image.

[0114] Each module in the aforementioned dark-field droplet image recognition device for nucleic acid molecules can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0115] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database can be used to store dark-field droplet images, as well as image data during processing. The network interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a method for dark-field droplet image recognition of nucleic acid molecules.

[0116] Those skilled in the art will understand thatFigure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0117] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the dark-field droplet image recognition method for nucleic acid molecules as described in any of the above embodiments.

[0118] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the dark-field droplet image recognition method for nucleic acid molecules as described in any of the above embodiments.

[0119] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the dark-field droplet image recognition method for nucleic acid molecules as described in any of the above embodiments.

[0120] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0121] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0122] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for dark-field droplet image recognition of nucleic acid molecules, characterized in that, The method includes: Acquire the dark field droplet image; calculate the segmentation thresholds for foreground and background markers in the dark field droplet image based on the dark field droplet image; use the Otsu thresholding method to segment the dark field droplet image according to the segmentation thresholds to obtain a binary image of the foreground in the dark field droplet image; The binary image of the foreground is eroded, and the erosion result is dilated to obtain the processed binary image of the foreground. Determine the maximum region in the foreground binary image, calculate the area of ​​the maximum region, and remove regions whose area in the foreground binary image is greater than a preset difference from the area of ​​the maximum region to obtain the foreground marker; For each first pixel in the foreground binary image, the distance between it and the second pixel that is closest to the first pixel is used to obtain a pixel image represented by distance; The watershed algorithm is used to determine the boundary line in the pixel image, and the background marker is extracted from the dark field droplet image based on the boundary line. The dark field droplet image is processed by adaptive H-minima transform, and the gradient magnitude of the dark field droplet image is obtained by using the Sobel operator to calculate the processed dark field droplet image. The foreground marker, background marker, and gradient magnitude of the image are superimposed to obtain a gradient magnitude image; Based on the gradient amplitude image, watershed segmentation is performed to obtain images of each droplet in the dark field droplet image, which are used for absolute quantitative analysis of nucleic acid molecules.

2. The method according to claim 1, characterized in that, Before acquiring the dark-field droplet image, the process includes: Obtain the original dark-field droplet image and perform Gaussian filtering on the original dark-field droplet image.

3. The method according to claim 2, characterized in that, The process of obtaining the original dark-field droplet image and then performing Gaussian filtering on the original dark-field droplet image includes: The original dark field droplet image after Gaussian filtering is enhanced to obtain the dark field droplet image.

4. The method according to claim 1, characterized in that, The step of calculating the segmentation thresholds for foreground and background markers in the dark-field droplet image based on the dark-field droplet image includes: The dark field droplet image is converted into a grayscale image, and then each grayscale value in the grayscale image is used as a threshold to divide all grayscale values ​​into two parts; Calculate the two grayscale values ​​corresponding to each grayscale value, and calculate the proportion of each part, the average grayscale value of each part, and the overall average grayscale value. Based on the proportion of each part of each gray value, the average gray value of each part, and the overall average gray value, the variance of the gray value is calculated, and the gray value with the largest variance is determined as the segmentation threshold.

5. A dark-field droplet image recognition device for nucleic acid molecules, characterized in that, The device includes: An extraction module is used to acquire the dark-field droplet image; calculate segmentation thresholds for foreground and background markers in the dark-field droplet image based on the image, and segment the dark-field droplet image using the Otsu thresholding method according to the segmentation thresholds to obtain a binary image of the foreground in the dark-field droplet image; erode the binary image of the foreground and dilate the erosion result to obtain a processed binary image of the foreground; determine the maxima region in the binary image of the foreground, calculate the area of ​​the maxima region, and compare the area of ​​the region in the binary image with the maxima region. Regions whose area difference is greater than a preset difference are removed to obtain the foreground marker; the distance between each first pixel in the foreground binary image and the second pixel closest to the first pixel is calculated to obtain a pixel image represented by distance; the watershed algorithm is used to determine the boundary line in the pixel image, and the background marker is extracted from the dark field droplet image based on the boundary line; the dark field droplet image is processed by adaptive H-minima transform, and the Sobel operator is used to calculate the gradient magnitude of the dark field droplet image; The overlay module is used to overlay the foreground marker, background marker, and gradient magnitude of the image to obtain a gradient magnitude image; The segmentation module is used to perform watershed segmentation based on the gradient amplitude image to obtain images of each droplet in the dark field droplet image, which is used for absolute quantitative analysis of nucleic acid molecules.

6. The apparatus according to claim 5, characterized in that, The device also includes a Gaussian filtering module for obtaining the original dark-field droplet image and performing Gaussian filtering on the original dark-field droplet image.

7. The apparatus according to claim 6, characterized in that, The device also includes a brightness enhancement processing module, used to enhance the brightness of the original dark field droplet image after Gaussian filtering to obtain the dark field droplet image.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images