Image processing method and system for measuring particle size of microfluidic liquid drop

Abstract

The invention relates to an image processing method and system for measuring the particle size of a micro-fluidic droplet, and the method comprises the steps: carrying out the convolution operation and smoothing of a Gaussian filter and a micro-droplet image, so as to remove image noise, and improve the precision of circle detection; carrying out binarization processing on the micro-droplet image so as to convert the micro-droplet image into a binary image with only two colors, thereby facilitating analysis and processing of the droplet; an interactive interface is provided for a user, a zooming instruction and a translation instruction input by the user on the interactive interface are received, personalized adjustment of the image size is realized, and a region of interest can be determined according to user preference to realize targeted observation; after the circles are detected, the goodness of fit of all the circles is calculated, and the optimal circle can be accurately screened according to circle parameters input by a user in an interactive interface so as to adapt to observation of liquid drops of different sizes; and finally, a visual image is formed based on a goodness-of-fit marking circle to visually present the form of the micro-droplet, so that an experimenter can perform rapid and deep droplet analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of microfluidic data processing, and particularly to an image processing method and system for measuring the particle size of microfluidic droplets. Background Art

[0002] In the observation of micro-droplets in a microfluidic chip, particle size measurement has important applications in multiple scientific and engineering fields, including materials science, pharmaceuticals, bioengineering, environmental science, and microfluidics. Accurately measuring the particle size distribution is crucial for understanding and controlling the physical and chemical properties of materials, optimizing production processes, and ensuring product quality.

[0003] Traditional particle size measurement methods mainly rely on instruments such as laser particle size analyzers and nano particle size analyzers to achieve, and have the following deficiencies: The sample needs to be a large-volume emulsion, such as 10 ml or 20 ml. The speed of gradually forming an emulsion by preparing micro-droplets through a microfluidic chip is very slow, which affects the real-time nature of the experiment and reduces the overall experimental efficiency; The equipment prices of laser particle size analyzers and nano particle size analyzers are expensive, and the maintenance costs are high, which is not conducive to popularization and application. Therefore, in the research field, image processing methods for measuring the particle size of micro-droplets have gradually emerged. However, in general image processing methods, only the particle size of a single or a small number of droplets can be measured at a time, the measurement efficiency is very low, and the content and format of the output observation data are relatively fixed, which is not conducive to the experimenter to conduct multi-angle and in-depth observations on micro-droplets. Summary of the Invention

[0004] Based on this, the purpose of the present invention is to provide an image processing method and system for measuring the particle size of microfluidic droplets, which have the advantages of low sample volume and high measurement efficiency. Moreover, the rich human-computer interaction function enables users to select the droplet regions of interest for targeted observations according to their needs, and form visual images to intuitively present the morphology of micro-droplets from multiple angles, which helps the experimenter to conduct rapid and in-depth particle size analysis.

[0005] The present invention is achieved through the following technical solutions:

[0006] According to the first aspect of the present application, an image processing method for measuring the particle size of microfluidic droplets is provided, including the following steps:

[0007] Obtain a micro-droplet image;

[0008] Perform a convolution operation on the Gaussian filter and the micro-droplet image using a convolution function to smooth the micro-droplet image, and perform binarization processing on the smoothed micro-droplet image to obtain a binarized image;

[0009] Receive the zoom instruction and pan instruction input by the user on the interaction interface, adjust the binarized image to a preset size, and then determine the sub-region to be measured;

[0010] Use the Hough transform algorithm to identify circles in the sub-region to be measured, and calculate the goodness of fit of each circle;

[0011] Receive the circle parameters input by the user in the interactive interface, screen the best circles in the binary image according to the circle parameters, mark the centers and goodness of fit of each best circle, and visually display them in the interactive interface.

[0012] In an optional embodiment, the convolution function is used to convolve the Gaussian filter with the micro-droplet image to smooth the micro-droplet image, and the smoothed micro-droplet image is binarized to obtain a binary image, including:

[0013] According to the size and standard deviation of the preset Gaussian kernel, a Gaussian filter is constructed based on the selected Gaussian function;

[0014] Receive the mask radius input by the user in the interactive interface, adjust the standard deviation of the Gaussian filter according to the mask radius, and calculate the coefficient matrix of the Gaussian filter according to the Gaussian function;

[0015] Use the coefficient matrix of the Gaussian filter as the input parameter of the convolution function, and use the convolution function to convolve the coefficient matrix of the Gaussian filter with the micro-droplet image to obtain the smoothed micro-droplet image;

[0016] Use an edge detection algorithm to detect the edges of each micro-droplet in the micro-droplet image;

[0017] Receive the binarization threshold input by the user in the interactive interface, and binarize the edge-detected micro-droplet image according to the binarization threshold to obtain a binary image of gray scale.

[0018] In an optional embodiment, the method for determining the binarization algorithm is: after performing convolution operation on the micro-droplet image, calculate the mean and standard deviation of the convolution image, and determine the binarization algorithm according to the mean and standard deviation of the convolution image.

[0019] In an optional embodiment, the circle parameters include a preset radius range, a preset edge thickness, a preset goodness of fit value, and a preset number of circles.

[0020] In an optional embodiment, the step of using the Hough transform algorithm to identify circles in the sub-region to be measured and calculate the goodness of fit of each circle includes:

[0021] Use a circle detection function to identify circles in the sub-region to be measured;

[0022] Receive the fitting parameters input by the user, and calculate the goodness of fit of each circle based on the fitting parameters and the goodness-of-fit calculation function.

[0023] In an alternative embodiment, the obtaining of the micro-droplet image includes:

[0024] Receive the image selection instruction input by the user on the interaction interface, and automatically identify and load the micro-droplet image that conforms to the preset format. Among them, the image selection instruction includes the selection instruction for a single micro-droplet image and / or the selection instruction for an image directory, and multiple micro-droplet images are included in the image directory.

[0025] In an alternative embodiment, it further includes the steps of: generating a fitting quality map according to the goodness of fit of each circle; and / or generating a circle radius histogram according to the radius of each circle.

[0026] In an alternative embodiment, based on different micro-droplet images, multiple micro-droplet detection processes are created to realize the synchronous detection and display of various types of droplets.

[0027] In an alternative embodiment, verify the result after visual display, export the final data of the visual result that meets the preset requirements, and generate a detection report.

[0028] According to the second aspect of the present application, there is provided an image processing system for measuring the particle size of microfluidic droplets. The image processing method for measuring the particle size of microfluidic droplets as described in the above embodiment is adopted to measure the particle size of micro-droplets and realize the visual observation of the micro-droplet distribution.

[0029] Compared with the prior art, the image processing method for measuring the particle size of microfluidic droplets in the embodiment of the present invention performs convolution operation on the Gaussian filter and the micro-droplet image through a convolution function to realize the smoothing process of the micro-droplet image and remove the noise of the micro-droplet image, so as to improve the accuracy of circle detection; perform binarization processing on the micro-droplet image to convert the micro-droplet image into a binary image with only two colors (usually black and white), so as to facilitate the analysis and processing of droplets; provide an interaction interface for the user to receive the zoom instruction and pan instruction input by the user on the interaction interface, realize the personalized adjustment of the image size, and can determine the region of interest according to the user's preference to achieve targeted observation; calculate the goodness of fit of each circle after detecting the circle. When screening the best circle that meets the preset standard, it can be accurately screened according to the circle parameters input by the user on the interaction interface to adapt to the observation of different sizes of droplets; finally, label the selected circles based on the goodness of fit to form a visual image to intuitively present the micro-droplet morphology from multiple angles, which helps the experimenter to perform rapid and in-depth particle size analysis and is conducive to the popularization and application of microfluidic technology.

[0030] For better understanding and implementation, the present invention will be described in detail below with reference to the accompanying drawings. Description of the Drawings

[0031] Figure 1 It is a flowchart of the steps of an image processing method for measuring the particle size of microfluidic droplets provided by an embodiment of the present invention;

[0032] Figure 2 It is a schematic diagram of the process of an image processing method for measuring the particle size of microfluidic droplets provided by an embodiment of the present invention;

[0033] Figure 3 It is a schematic diagram of the image recognition operation during the image loading process provided by an embodiment of the present invention;

[0034] Figure 4 It is a schematic diagram of image convolution processing and image binarization provided by an embodiment of the present invention;

[0035] Figure 5 It is a schematic diagram of image scaling and sub-region selection after binarization processing provided by an embodiment of the present invention;

[0036] Figure 6 It is a schematic diagram of image annotation, fitting quality diagram, and circular radius histogram provided by an embodiment of the present invention;

[0037] Figure 7 It is an observation diagram of the inner radius of a circular droplet provided by an embodiment of the present invention;

[0038] Figure 8 It is an observation diagram of the middle radius of a circular droplet provided by an embodiment of the present invention;

[0039] Figure 9 It is an observation diagram of the outer radius of a circular droplet provided by an embodiment of the present invention;

[0040] Figure 10 It is an observation diagram of the edge thickness of a circular droplet provided by an embodiment of the present invention;

[0041] Figure 11 It is an observation diagram of the sorting of circular droplets provided by an embodiment of the present invention;

[0042] Figure 12 It is an observation diagram of the contour of a circular droplet provided by an embodiment of the present invention; Detailed Embodiments

[0043] It should be clear that the described embodiments are only some embodiments of the embodiments of the present application, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope protected by the embodiments of the present application.

[0044] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "center", "longitudinal", "transverse", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In the description of the present invention, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0045] It should be noted that when an element is referred to as "fixed to" another element, it can be directly on the other element or there may also be an intermediate element. When an element is considered to be "connected" or "fixedly connected" to another element, it can be directly connected to the other element or there may be an intermediate element at the same time.

[0046] In the field of microfluidics technology, when using instruments such as laser particle size analyzers and nano particle size analyzers to measure the droplet size, there are problems such as a large sample volume, slow detection speed, low detection efficiency, and low measurement accuracy. When using general image processing methods to detect the size of micro-droplets, only the size of a single or a small number of droplets can be measured at a time, and the measurement efficiency is still very low. Moreover, the content and format of the output observation data are relatively fixed, which is not conducive to the experimenter's multi-angle and in-depth observation of micro-droplets and the popularization and application of microfluidics technology.

[0047] Based on this, the embodiments of the present application provide an image processing method and system for measuring the size of microfluidic droplets, which have the advantages of low sample volume and high measurement efficiency. Moreover, the rich human-computer interaction function enables users to select the droplet area of interest for targeted observation according to their needs, and form a visual image to intuitively present the morphology of micro-droplets from multiple angles, which helps the experimenter to perform rapid and in-depth particle size analysis and is conducive to the popularization and application of microfluidics technology.

[0048] As Figure 1 、 Figure 2 shown, the image processing method for measuring the size of microfluidic droplets includes the following steps:

[0049] S1: Obtain a micro-droplet image;

[0050] S2: Perform a convolution operation on the Gaussian filter and the micro-droplet image using a convolution function to smooth the micro-droplet image, and perform binarization processing on the smoothed micro-droplet image to obtain a binarized image;

[0051] S3: Receive the zoom instruction and pan instruction input by the user on the interaction interface, adjust the binarized image to a preset size, and then determine the sub-region to be measured;

[0052] S4: Identify circles in the sub-region to be measured using the Hough transform algorithm, and calculate the goodness of fit of each circle.

[0053] S5: Receive the circle parameters input by the user in the interactive interface, filter out the best circles in the binary image according to the circle parameters, mark the centers and goodness of fit of each best circle, and visually display them in the interactive interface.

[0054] In this embodiment, the image processing method for measuring the particle size of microfluidic droplets performs a convolution operation on the Gaussian filter and the micro-droplet image through a convolution function to smooth the micro-droplet image, remove the noise of the micro-droplet image, so as to improve the accuracy of circle detection; perform binary processing on the micro-droplet image to convert the micro-droplet image into a binary image with only two colors (usually black and white), so as to facilitate the analysis and processing of droplets; provide an interactive interface for users, receive the zoom instruction and pan instruction input by the user in the interactive interface, realize personalized adjustment of the image size, and can determine the region of interest according to the user's preference to achieve targeted observation; calculate the goodness of fit of each circle after detecting the circles, and when screening the best circles that meet the preset criteria, it can be accurately screened according to the circle parameters input by the user in the interactive interface to adapt to the observation of droplets of different sizes; finally, mark the selected circles based on the goodness of fit to form a visual image to intuitively present the micro-droplet morphology from multiple angles, which helps the experimenter to perform rapid and in-depth particle size analysis and is conducive to the popularization and application of microfluidic technology.

[0055] In step S1, obtaining the micro-droplet image includes: receiving the image selection instruction input by the user in the interactive interface, automatically identifying and loading the micro-droplet image that conforms to the preset format, where the image selection instruction includes the selection instruction for a single micro-droplet image and / or the selection instruction for an image directory, and the image directory includes multiple micro-droplet images.

[0056] As Figure 3 shown, the image processing system is set with a human-computer interaction function, displays the interactive interface, and the user can input instructions in the interactive interface to perform relevant operations. When selecting an image, the user can conveniently select a single image or the entire folder for processing. The preset format of the micro-droplet image can include.jpg,.tif,.bmp, etc., thus eliminating the separate image format conversion step and improving the image processing efficiency. Automatically identify and load all images that conform to the format in the user-specified directory, support batch processing of multiple samples, and can also improve the experimental efficiency.

[0057] In step S2, as Figure 4As shown, a convolution function is used to perform a convolution operation on the Gaussian filter and the micro-droplet image to smooth the micro-droplet image, and the smoothed micro-droplet image is binarized to obtain a binary image, including the following sub-steps:

[0058] S21: Based on the preset size and standard deviation of the Gaussian kernel, a Gaussian filter is constructed based on the selected Gaussian function.

[0059] In this step, the Gaussian function can be where (x, y) are the coordinates of the pixel point in the mask relative to the center, ο is the standard deviation, usually related to the mask radius, generally σ = r can be taken, or adjusted according to the specific image characteristics and processing requirements.

[0060] S22: Accept the mask radius input by the user in the interactive interface, adjust the standard deviation of the Gaussian filter according to the mask radius, and calculate the coefficient matrix of the Gaussian filter according to the Gaussian function.

[0061] In this step, after calculating the values of the two-dimensional Gaussian function at each coordinate (x, y), these values are combined into a two-dimensional matrix, which is the coefficient matrix of the Gaussian filter. To ensure that the overall brightness of the image will not change during the convolution operation, the coefficient matrix is also normalized so that the sum of the coefficients is 1.

[0062] It should be noted that the larger the mask radius, the more pixels in the neighborhood will be considered during the convolution operation, resulting in a stronger smoothing effect. A larger mask radius can more effectively remove noise in the image, but it may also over-smooth the edges, resulting in a decrease in the edge position and clarity; on the contrary, a smaller mask radius considers fewer neighborhood pixels, has a weaker smoothing effect, and can more accurately retain edge information because it only smooths a small number of pixels near the edge. Therefore, the user can observe the processing effect of the micro-droplet image in real time and continuously adjust the mask radius according to the processing effect to achieve a satisfactory effect.

[0063] S23: Use the coefficient matrix of the Gaussian filter as the input parameter of the convolution function, and perform a convolution operation on the coefficient matrix of the Gaussian filter and the micro-droplet image using the convolution function to obtain the smoothed micro-droplet image.

[0064] In this step, the convolution function is the conv2 function. The coefficient matrix of the Gaussian filter is used as an input parameter of the conv2 function, and the micro-droplet image to be processed is used as another input parameter. The conv2 function will move the Gaussian filter point by point on the micro-droplet image. For each pixel in the micro-droplet image, the pixel values in the area with the same size as the Gaussian mask around it are multiplied corresponding to the coefficients of the Gaussian filter, and the product results are added to obtain the new value of the pixel after convolution. Traverse all the pixels of the micro-droplet image to complete the convolution operation of the entire micro-droplet image with the Gaussian filter, and obtain the smoothed micro-droplet image.

[0065] S24: Use an edge detection algorithm to perform edge detection on each micro-droplet in the micro-droplet image to enhance the circular feature of the micro-droplet.

[0066] In this embodiment, the Canny edge detection algorithm is used to perform edge detection on each micro-droplet in the micro-droplet image.

[0067] S25: Receive the binarization threshold input by the user on the interaction interface, and perform binarization processing on the micro-droplet image after edge detection according to the binarization threshold to obtain a grayscale binarized image.

[0068] The binarization threshold can clearly distinguish the droplet from the background. The experimenter inputs the binarization threshold on the interaction interface according to experimental experience. If the input binarization threshold is inappropriate, it can be adjusted while observing to determine the optimal binarization threshold, so as to divide the image into two categories: foreground and background. According to the determined binarization threshold, the grayscale value of each pixel in the image is compared with the binarization threshold. If the grayscale value of the pixel is greater than the binarization threshold, it is set to white (usually represented by 255), indicating the droplet area; if the grayscale value of the pixel is less than or equal to the binarization threshold, it is set to black (usually represented by 0), indicating the background area, and thus a preliminary binarized image is obtained.

[0069] In an optional embodiment, the determination method of the binarization algorithm is: after performing a convolution operation on the micro-droplet image, calculate the mean and standard deviation of the convolution image, and determine the binarization algorithm according to the mean and standard deviation of the convolution image.

[0070] In step S3, as Figure 5 shown, receive the zoom instruction and translation instruction input by the user on the interaction interface, adjust the binarized image to the preset size and then determine the sub-region to be measured, so as to view the region of interest according to the user's preference and improve the observation flexibility.

[0071] In step S4, use the Hough transform algorithm to identify the circles in the sub-region to be measured, and calculate the goodness of fit of each circle, including the following sub-steps:

[0072] S41: Identify circles in the sub-region to be measured using a circle detection function;

[0073] S42: Receive the fitting parameters input by the user, and calculate the goodness of fit of each circle based on the fitting parameters and the goodness-of-fit calculation function.

[0074] In this step, the circle detection function is the cv2.HoughCircles() function. The goodness of fit is used to measure the matching degree between the detected circle and the actual circle in the image. For each detected circle, the goodness-of-fit calculation function can be called to calculate its goodness of fit based on the fitting parameters input by the user.

[0075] In step S5, receive the circle parameters input by the user in the interactive interface, screen out the best circles that meet the preset criteria in the binary image, mark the centers and goodness of fit of each best circle, and visually display them in the interactive interface. Among them, the circle parameters include the preset radius range, the preset edge thickness, the preset goodness-of-fit value, and the preset number of circles.

[0076] The user can set the radius range for identifying circles to adapt to the detection of droplets of different sizes; control the edge thickness of the identified circles, which affects the fitting quality; screen out the circles that meet the requirements according to the size of the goodness of fit, for example, only keep the circles with a goodness of fit greater than a certain threshold, set the number of preset circles, and screen out a certain number of best circles for in-depth observation.

[0077] As Figure 6 shown, in the visualization method, the identified center points can be marked on the original micro-droplet image, and the fitting quality can be represented by colors, with red indicating poor fitting quality and blue indicating good fitting quality.

[0078] Optionally, the visualization method can also include generating a merit plot and a radius histogram of circles. Therefore, the image processing method for measuring the particle size of microfluidic droplets further includes the steps of: generating a merit plot according to the goodness of fit of each circle; and / or generating a radius histogram of circles according to the radius of each circle.

[0079] The merit plot is used to show the relationship between the goodness of fit and the found circles, helping the user evaluate the reliability of the recognition results. As shown in Figure..., in the merit plot, the horizontal axis is used to represent the numbers of the selected circles, and the vertical axis is used to represent the quality of the fitting effect, with red indicating poor fitting quality and blue indicating good fitting quality.

[0080] The radius histogram of circles is used to show the radius distribution of the identified circles, and supports display in pixels or calibration units. As shown in Figure..., in the radius histogram of circles, the horizontal axis is used to represent the radius distribution of the circles, and the vertical axis is used to represent the number of circles with a certain radius.

[0081] In an optional embodiment, the image processing method may further include the steps of: based on different micro-droplet images, creating multiple micro-droplet detection processes to achieve synchronous detection and display of multiple types of droplets, so as to synchronously achieve comparative observation of multiple types of micro-droplets, further improving the flexibility of the image processing system and the batch processing efficiency.

[0082] In an optional embodiment, it further includes the steps of: verifying the visualized result, and exporting the final data of the visualized result that meets the preset requirements to generate a detection report.

[0083] Specifically, when the user clicks the "Export Data" button, the image processing system exports the currently displayed circular data as a structure for further data analysis and processing in MATLAB. Then the user can also reduce the number of displayed circles through a slider to automatically remove circles with poor fitting quality, ensuring the quality of the exported data. When the working directory contains multiple image files, the image processing system automatically processes each image and displays the results in different windows, supporting efficient batch analysis.

[0084] Optionally, as Figures 7 to 12 shown, the user can also select the droplet features of interest on the interaction interface for in-depth observation, including the inner radius observation of circular droplets, the middle radius observation of circular droplets, the outer radius observation of circular droplets, the edge thickness observation of circular droplets, the sorting observation of circular droplets, and the contour observation of circular droplets, so as to achieve multi-angle intuitive observation of micro-droplets, which is helpful for the experimenter to conduct in-depth analysis of micro-droplets and is helpful for the popularization and application of microfluidic technology.

[0085] Compared with the prior art, the image processing method for measuring the particle size of microfluidic droplets of the present invention performs a convolution operation on the Gaussian filter and the micro-droplet image through a convolution function to achieve smoothing processing of the micro-droplet image and remove the noise of the micro-droplet image, so as to improve the accuracy of circle detection; performs binarization processing on the micro-droplet image to convert the micro-droplet image into a binary image with only two colors (usually black and white) for easy analysis and processing of the droplets; provides an interaction interface for the user to receive the zoom instruction and translation instruction input by the user on the interaction interface, realizes personalized adjustment of the image size, and can determine the region of interest according to the user's preference for targeted observation; calculates the goodness of fit of each circle after detecting the circles, and when screening the best circles that meet the preset criteria, can accurately screen according to the circle parameters input by the user on the interaction interface to adapt to the observation of different-sized droplets; finally, annotates the selected circles based on the goodness of fit to form a visualized image to intuitively present the micro-droplet morphology from multiple angles.

[0086] Overall, the image processing method has the advantages of low sample volume and high measurement efficiency. Moreover, its rich human-computer interaction functions enable users to select droplet regions of interest for targeted observation according to their needs, and form visual images to intuitively present the morphology of micro-droplets from multiple angles, which helps experimenters to conduct rapid and in-depth particle size analysis and is conducive to the popularization and application of microfluidic technology.

[0087] An embodiment of the present application further provides an image processing system for measuring the particle size of microfluidic droplets. Based on a computer, it uses the image processing method for measuring the particle size of microfluidic droplets as described in the above embodiment to measure the particle size of micro-droplets and realize visual observation of the micro-droplet distribution.

[0088] The image processing system and the above image processing method are two embodiments under the same inventive concept. The content not described in this embodiment can refer to the specific content of the image processing method. They can apply the same technical solution, solve the same technical problem, and achieve the same technical effect, so it will not be elaborated in this embodiment.

[0089] The above embodiments only represent several implementation manners of the present invention. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the inventive concept of the present invention, several modifications and improvements can still be made, and the present invention also intends to include these changes and modifications.

Claims

1. An image processing method for measuring the particle size of microfluidic droplets, characterized in that: The following steps are involved: Acquire microdroplet images; Using a convolution function to perform a convolution operation on a Gaussian filter and the micro-droplet image to smooth the micro-droplet image, and binarizing the smoothed micro-droplet image to obtain a binary image; Receiving a zoom instruction and a translation instruction input by a user in an interactive interface, and determining a sub-region to be measured after adjusting the binary image to a preset size; Using the Hough transform algorithm to identify the circles in the sub-region to be measured, and calculating the goodness of fit of each circle; The circle parameters input by the user in the interactive interface are received, the best circle is selected in the binary image according to the circle parameters, the center and goodness of fit of each best circle are marked, and the circles are visually displayed in the interactive interface.

2. The image processing method for measuring the particle size of microfluidic droplets according to claim 1, characterized in that: The step of using a convolution function to perform a convolution operation on the Gaussian filter and the micro-droplet image to smooth the micro-droplet image, and binarizing the smoothed micro-droplet image to obtain a binarized image comprises: According to the size and standard deviation of the preset Gaussian kernel, a Gaussian filter is constructed based on the selected Gaussian function; Accepting a mask radius input by a user in an interactive interface, adjusting a standard deviation of the Gaussian filter according to the mask radius, and calculating a coefficient matrix of the Gaussian filter according to a Gaussian function; Using the coefficient matrix of the Gaussian filter as an input parameter of a convolution function, and using the convolution function to perform a convolution operation on the coefficient matrix of the Gaussian filter and the micro-droplet image to obtain the micro-droplet image after smoothing; Performing edge detection on each micro-droplet in the micro-droplet image using an edge detection algorithm; A binarization threshold value inputted by a user in an interactive interface is received, and the micro-droplet image after edge detection is binarized according to the binarization threshold value to obtain a grayscale binarized image.

3. The image processing method for measuring the particle size of microfluidic droplets according to claim 2, characterized in that: The method for determining the binarization algorithm is as follows: after performing a convolution operation on the micro-droplet image, the mean and standard deviation of the convolution image are calculated, and the binarization algorithm is determined according to the mean and standard deviation of the convolution image.

4. The image processing method for measuring the particle size of microfluidic droplets according to claim 2, characterized in that: The circle parameters include a preset radius range, a preset edge thickness, a preset goodness of fit value, and a preset number of circles.

5. The image processing method for measuring the particle size of microfluidic droplets according to claim 1, characterized in that: The method of using the Hough transform algorithm to identify the circles in the sub-region to be measured and calculating the goodness of fit of each circle includes: Using a circle detection function to identify a circle in the sub-region to be detected; The fitting parameters input by the user are received, and the fitting goodness of each circle is calculated based on the fitting parameters and the fitting goodness calculation function.

6. The image processing method for measuring the particle size of microfluidic droplets according to claim 1, characterized in that: The step of acquiring the micro-droplet image comprises: Receive an image selection instruction input by a user in an interactive interface, automatically identify and load micro-droplet images that conform to a preset format, wherein the image selection instruction includes a selection instruction for a single micro-droplet image, and / or a selection instruction for an image directory, wherein the image directory includes multiple micro-droplet images.

7. The image processing method for measuring the particle size of microfluidic droplets according to claim 1, characterized in that: Also includes the steps: generating a quality of fit plot based on the goodness of fit of each circle; and / or, Generate a circle radius histogram based on the radius of each circle.

8. The image processing method for measuring the particle size of microfluidic droplets according to claim 1, characterized in that: Based on different micro-droplet images, multiple micro-droplet detection processes are created to achieve synchronous detection and display of multiple types of droplets.

9. The image processing method for measuring the particle size of microfluidic droplets according to claim 1, characterized in that: Verify the results after visualization, export the final data of the visualization results that meet the preset requirements, and generate a test report.

10. An image processing system for measuring the particle size of microfluidic droplets, characterized in that: The image processing method for measuring the particle size of microfluidic droplets as claimed in any one of claims 1 to 9 is used to measure the particle size of microdroplets and realize visual observation of microdroplet distribution.

Images