Image processing method and processing device for evaluating injection effect of needle-free syringe

Abstract

The present invention relates to an image processing method and a processing device for evaluating the injection effect of a needle-free syringe. The image contains a measurement object generated by a needle-free syringe test, and the method comprises the following steps: preprocessing the image captured by a camera to remove irrelevant parts of the image and highlight key information in the image by image enhancement; extracting the measurement object from the preprocessed image by a combination of threshold segmentation and region segmentation, and using a corresponding threshold segmentation method to obtain the measurement object according to the type of the image; measuring the shape information of the extracted measurement object, and using a corresponding shape information measurement method to obtain at least a part of the shape information of the measurement object according to the type of the image. The present invention can realize quantitative evaluation of the injection effect of a needle-free syringe.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to an image processing method and an image processing device for evaluating the injection effect of a needle-free syringe. Background Art

[0002] At present, the evaluation of the injection effect of needle-free syringes mainly relies on the quantitative consideration of four indicators: the diameter of the skin papule, the depth of drug diffusion, the diameter of drug diffusion and the length of the penetration hole. In addition to direct animal experiments, polyacrylamide gel is also used as a model to simulate the skin. The current measurement method is mainly to use a vernier caliper, which is affected by subjective factors, inefficient and cannot be verified.

[0003] Image processing refers to the use of algorithms and codes on computers to automatically process, manipulate, analyze and interpret images. It is widely used in many disciplines and fields, such as television, photography, robotics, remote sensing, medical diagnosis and industrial inspection. Image processing methods extract image data, preprocess, enhance, restore, represent, segment, classify, detect and identify (objects) to better analyze, understand and interpret data. Therefore, the above indicators can be obtained based on image processing, so that the injection effect of needle-free syringes can be evaluated.

[0004] The above description of the background technology is only for facilitating an in-depth understanding of the technical solution of the present invention (such as the technical means used, the technical problems solved and the technical effects produced), and should not be regarded as an admission or any form of implication that the information constitutes prior art already known to those skilled in the art. Summary of the invention

[0005] The present invention provides an image processing method and a processing device for evaluating the injection effect of a needle-free syringe, which can realize quantitative evaluation of the injection effect of the needle-free syringe.

[0006] According to one embodiment of the present invention, there is provided a method for processing an image for evaluating the injection effect of a needle-free syringe, wherein the image has a measurement object generated by a needle-free syringe test, and the method for processing an image for evaluating the injection effect of a needle-free syringe comprises the following steps: preprocessing the image captured by a camera to remove irrelevant parts of the image and highlighting key information in the image by image enhancement; extracting the measurement object from the preprocessed image by a combination of threshold segmentation and region segmentation, and obtaining the measurement object by using a corresponding threshold segmentation method according to the type of the image; measuring shape information of the extracted measurement object, and obtaining at least a portion of shape information of the measurement object by using a corresponding shape information measurement method according to the type of the image.

[0007] The images can be classified into a first type of image, a second type of image, and a third type of image, and the edge clarity of the measuring object in the first type of image is stronger than that in the second type of image and the third type of image, all shape information of the measuring object in the third type of image does not require extraction of the edge of the measuring object during measurement, while at least a portion of the shape information of the measuring object in the first type of image and the second type of image requires extraction of the edge of the measuring object during measurement.

[0008] The steps of using the corresponding threshold segmentation method according to the type of image may include: determining whether the image is a first type of image; when it is determined that the image is a first type of image, performing threshold segmentation using an iterative threshold method; when it is determined that the image is not a first type of image, performing threshold segmentation using a maximum entropy threshold segmentation method.

[0009] After performing threshold segmentation on the image, the region growing method can be used to perform region segmentation on the image.

[0010] The steps of using a corresponding shape information measurement method according to the type of the image may include: determining whether the image is a third type image; when it is determined that the image is a third type image, directly measuring the shape information of the extracted measurement object; when it is determined that the image is not a third type image, extracting the edge of the measurement object, receiving a selection of a measurement start point and a measurement end point, and at least one of the measurement start point and the measurement end point is on the edge of the measurement object, and measuring the distance between the measurement start point and the measurement end point to measure at least a portion of the shape information of the measurement object.

[0011] The step of directly measuring the extracted shape information of the measurement object may include: calling a regionprops function to measure the shape information of the measurement object.

[0012] The step of extracting the edge of the measured object may include: smoothing the edge of the measured object, and extracting the edge of the measured object using a canny edge extraction method.

[0013] The step of receiving selections of a measurement start point and a measurement end point may include: calling a distance drawing function, receiving selection of a measurement start point by a user clicking a mouse, and receiving selection of a measurement end point by a user clicking a mouse again.

[0014] The image may further include a reference object of known size, and the step of measuring the distance between the measurement start point and the measurement end point includes: receiving a modification to the measurement label scale, and calculating the actual distance between the measurement start point and the measurement end point according to the modified measurement label scale.

[0015] The first type of image is an image of injecting a drug solution into a substance used to simulate skin; the second type of image is a cross-sectional image of the skin after injecting the drug solution into the animal's skin; and the third type of image is an image of a skin bump generated after injecting the drug solution into the animal's skin.

[0016] All shape information of the measured object in the first type of image includes: the length and width of the minimum circumscribed rectangle of the measured object, the length from the upper end point of the measured object to a specific point inside the measured object along the length direction of the measured object, the length of the minimum circumscribed rectangle of the measured object indicates the diffusion depth of the drug solution, the width of the minimum circumscribed rectangle of the measured object indicates the diffusion diameter of the drug solution, and the length from the upper end point of the measured object to a specific point inside the measured object along the length direction of the measured object indicates the depth of the hole; all shape information of the measured object in the second type of image includes: the longest distance of the measured object in the transverse direction and the longest distance in the longitudinal direction, the longest distance of the measured object in the transverse direction indicates the diffusion width of the intradermal injection, and the longest distance of the measured object in the longitudinal direction indicates the diffusion depth of the intradermal injection; all shape information of the measured object in the third type of image includes the diameter of the measured object, and the diameter of the measured object indicates the diameter of the skin papule.

[0017] The steps of image preprocessing may include: receiving an image captured by a camera and displaying the image; cropping the image; performing RGB separation on the cropped image; and enhancing the image by adjusting the contrast and / or grayscale value of the separated G channel image or R channel image.

[0018] The step of performing RGB separation on the cropped image may include: determining whether the liquid medicine injected during the needle-free syringe test is rhodamine B dye; when it is determined that the liquid medicine injected during the needle-free syringe test is rhodamine B dye, separating the cropped image into a G channel image; when it is determined that the liquid medicine injected during the needle-free syringe test is not rhodamine B dye, separating the cropped image into an R channel image.

[0019] According to another embodiment of the present invention, there is provided a device for processing images for evaluating the injection effect of a needle-free syringe, which includes: at least one data processor; and at least one memory storing instructions, which, when executed by the at least one data processor, executes the above-mentioned method for processing images for evaluating the injection effect of a needle-free syringe.

[0020] The present invention adopts the above technical scheme, which has the following beneficial effects: the present invention can automatically measure the shape information of the liquid medicine injected into the substance or animal body by processing the image after the needle-free syringe test. The injection effect of the needle-free syringe can be evaluated through this shape information, and then the quantitative evaluation of the injection effect of the needle-free syringe can be realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The following will describe the exemplary embodiments of the present invention in more detail with reference to the accompanying drawings. For the sake of clarity, the same components in different drawings are shown with the same reference numerals. It should be noted that the drawings are only for illustration and are not necessarily drawn to scale. In these drawings:

[0022] Figure 1A A first type of image among images used for evaluating the injection effect of a needle-free injector is exemplarily shown.

[0023] Figure 1B Shows Figure 1A The image after preprocessing.

[0024] Figure 1C Shows from Figure 1B The edge of the object to be measured is extracted.

[0025] Figure 1D The extracted edge of the measured object is shown Figure 1A The superimposed image.

[0026] Figure 2A The second type of images among the images used for evaluating the injection effect of the needle-free injector are exemplarily shown.

[0027] Figure 2B Shows Figure 2A The image after preprocessing.

[0028] Figure 2C Shows from Figure 2B The edge of the object to be measured is extracted.

[0029] Figure 2D The extracted edge of the measured object is shown Figure 2A The superimposed image.

[0030] Figure 3A The third type of images among the images used for evaluating the injection effect of a needle-free injector is exemplarily shown.

[0031] Figure 3B Shows Figure 3A The image after preprocessing.

[0032] Figure 3C shows the bounding box drawn with Figure 3A The superimposed image.

[0033] Figure 4 is a flowchart of a method for processing an image for evaluating an injection effect of a needle-free injector according to various embodiments of the present invention.

[0034] Figure 5 is a detailed flowchart of a method of processing an image for evaluating an injection effect of a needle-free injector according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0035] The implementation scheme of the present invention is described in detail below. This implementation scheme is implemented on the premise of the technical scheme of the present invention, and a detailed implementation method and specific operation process are given, but the protection scope of the present invention is not limited to the implementation scheme described below.

[0036] In an embodiment of the present invention, the image has measurement objects generated by a needle-free syringe test. Figure 1A , Figure 2A and Figure 3A Three types of images among images used for evaluating the injection effect of a needle-free syringe are exemplarily shown.

[0037] Figure 1A The first type of image is exemplarily shown, and the first type of image is an image of injecting a liquid medicine into a substance used to simulate skin. Figure 1A In the case shown, a needle-free syringe injects the dye Rhodamine B into a polyacrylamide gel used to simulate skin. Figure 1A The pink-colored area shown (ie, the Rhodamine B stain diffused in the polyacrylamide gel) is the object of measurement.

[0038] Figure 2A The second type of image is exemplarily shown, which is a cross-sectional image of the skin of an animal after a drug solution is injected into the skin of the animal body. The "animal body" may include vertebrates, but is not limited to farm or production animals (such as pigs, cattle, poultry, horses), pets, primates, mice and rats. Specifically, in Figure 2A In the described case, a needle-free syringe was used to inject methylene blue dye into the skin of the rats (i.e., into the epidermis and dermis of the skin). Figure 2A The blue area shown (ie, the methylene blue dye diffused in the rat skin) is the object of measurement.

[0039] Figure 3A The third type of image is exemplarily shown, and the third type of image is an image of a skin papule generated after injecting a drug solution into the skin of an animal. Figure 3A In the case shown, methylene blue dye was injected intradermally into rats using a needle-free syringe. Figure 3A The blue area (ie, the dermis) shown is the object of measurement.

[0040] In testing the injection effect of a needle-free syringe, the measuring object (e.g. Figure 1A , Figure 2A and Figure 3AThe shape information of the measurement object shown in the figure can evaluate the injection effect of the needle-free syringe. For example, Figure 1A , Figure 2A and Figure 3A The length, width or diameter of the measurement object shown can be used to evaluate the injection effect of the needle-free injector.

[0041] Images (e.g. Figure 1A , Figure 2A and Figure 3A ) can be captured by a camera. Figure 4 is a flow chart of a method for processing an image for evaluating the injection effect of a needle-free injector according to various embodiments of the present invention. Figure 4 As shown, the image processing method for evaluating the injection effect of a needle-free syringe according to various embodiments of the present invention includes the following steps: preprocessing the image captured by the camera (S10) to remove irrelevant parts of the image and enhance the image. Extracting the measurement object from the preprocessed image by a combination of threshold segmentation and region segmentation (S20), and using the corresponding threshold segmentation method to obtain the measurement object according to the type of the image. Measuring the shape information of the extracted measurement object (S30), and using the corresponding shape information measurement method to obtain at least a part of the shape information of the measurement object according to the type of the image.

[0042] Figure 5 FIG. 4 is a detailed flowchart of a method for processing an image for evaluating the injection effect of a needle-free syringe according to an exemplary embodiment of the present invention. Figure 5 As shown, the step of pre-processing the image ( S10 ) may include receiving an image captured by a camera, and displaying the image ( S11 ).

[0043] Next, the image is cropped (S12) to remove irrelevant parts of the image. Specifically, the number of rows, columns and pixels of the image is determined, and a function (e.g., imcrop function) is called according to the required size to crop the image into a rectangular local area. For example, after cropping the image, the following can be obtained: Figure 1A , Figure 2A and Figure 3A Image shown.

[0044] Then, the cropped image is subjected to red, green and blue (RGB) channel separation. RGB channel separation refers to separating the image into three channels: red (R), green (G), and blue (B).

[0045] According to the principle of three primary colors, red is bright in the R channel and dark in the G and B channels; green is bright in the G channel and dark in the R and B channels; blue is bright in the B channel and dark in the R and G channels. Figure 1A , Figure 2A and Figure 3A It can be seen that different colors of dyes can be used for the test of needle-free syringes. Different colors of dyes cause different colors of the measurement objects in the image. If the same RGB channel separation standard is used, the measurement object and non-measurement object area in each image cannot be clearly distinguished.

[0046] Therefore, according to an embodiment of the present invention, the step of performing RGB separation on the cropped image may include: determining whether the liquid medicine injected during the needle-free syringe test is a rhodamine B dye (S13), and when it is determined that the liquid medicine injected during the needle-free syringe test is a rhodamine B dye ("Yes" of S13), the cropped image may be separated into a G channel image (S14). When it is determined that the liquid medicine injected during the needle-free syringe test is not a rhodamine B dye ("No" of S13), indicating that the liquid medicine injected by the needle-free syringe is a methylene blue dye, the cropped image may be separated into an R channel image (S15). Therefore, the measurement object may appear as a dark area in the separated G channel image or R channel image, and be clearly distinguished from the non-measurement object area.

[0047] Specifically, the category of the liquid medicine injected during the needle-free syringe test can be determined by identifying the artificial label carried in the image (the artificial label marks whether the liquid medicine in the image is a rhodamine B dye or a methylene blue dye). In other embodiments, the image can be input into a trained machine learning model to determine the category of the liquid medicine injected during the needle-free syringe test.

[0048] In addition, when the image is separated into a G channel image or an R channel image, the separated G channel image or R channel image is a grayscale image. In this case, the image can be enhanced (S16) by adjusting the contrast and / or grayscale value of the separated G channel image or R channel image, thereby highlighting the key information in the image. Specifically, a function (e.g., imadjust function) can be called first to enhance the contrast of the image, and then the image can be further enhanced by adjusting the grayscale value.

[0049] Figure 1B Shows Figure 1A The image after preprocessing. Figure 2B Shows Figure 2A The image after preprocessing. Figure 3B Shows Figure 3A After preprocessing, the image is compared Figure 1B and Figure 1A , Figure 2B and Figure 2A as well as Figure 3B and Figure 3AIt can be seen that the details in the original image are fuzzy and the brightness is uneven, but after preprocessing, the details in the image become significantly clearer, the brightness becomes uniform, the proportion of each gray level is more balanced, and the image quality is improved, which lays a good foundation for the subsequent image segmentation.

[0050] According to an embodiment of the present invention, the step of extracting the measurement object from the preprocessed image by combining threshold segmentation and region segmentation may include: first performing threshold segmentation and then performing region segmentation to extract the measurement object.

[0051] In addition, in the threshold segmentation, the corresponding threshold segmentation method is used according to the type of image. Specifically, back to Figure 5 After enhancing the image (S16), it is determined whether the image is a first type image (S21). When it is determined that the image is a first type image ("Yes" of S21), threshold segmentation is performed using an iterative threshold method (S22). When it is determined that the image is not a first type image ("No" of S21), that is, when it is determined that the image is a second type image or a third type image, threshold segmentation is performed using a maximum entropy threshold segmentation method (S23).

[0052] Specifically, the type of the image can be determined by identifying an artificial label carried in the image (the artificial label marks whether the image is a first type image, a second type image, or a third type image). In other embodiments, the image can be input into a trained machine learning model to determine the type of the image.

[0053] Among them, the first type of image (for example, Figure 1A ) has a clearer edge of the measured object than the second type of image (for example, Figure 2A ) and a third type of image (e.g. Figure 3A ) has the edge clarity of the measured object. For example, Figure 1A The edges of the dye injected into the gel are clear, while Figure 2A and Figure 3A The edges of the dye injected into the animal will smudge and spread into the skin tissue.

[0054] The basic assumption of the threshold segmentation method is that the image is composed of multiple sub-regions, and the multiple sub-regions are respectively composed of pixels with similar grayscale. The threshold segmentation method takes this as a starting point, and by selecting a suitable threshold, the pixels above the threshold are divided into one region, and the pixels below the threshold are classified into another region. The key to the threshold segmentation method lies in the selection of the threshold. There are many ways to select the threshold, for example, the iterative threshold method and the maximum entropy threshold segmentation method used in the embodiments of the present invention.

[0055] The basic principle of the iterative threshold method is: first select an initial threshold T0 (generally taking the average of the maximum and minimum values ​​of all grayscales), use this threshold to divide the pixels above the threshold into one group, and the pixels below the threshold into another group, so as to divide the image into two groups, and then take the average grayscale of the pixels in the two groups of images, and take the middle value of the average grayscale of the two groups of images as the new threshold T, and repeat the above process until the difference between the thresholds of the two iterations is not greater than a relatively small predefined parameter δ T .

[0056] The basic principle of the maximum entropy threshold segmentation method is: first, count the number of gray levels of each pixel in the image, calculate the probability value of each pixel gray level, and then calculate the entropy value with the largest product of the probability values, indicating that the information entropy is the largest, and the threshold with the largest entropy value is the optimal threshold.

[0057] Since the maximum entropy threshold segmentation method can better preserve the information of edges and weak objects, the second type of images and the third type of images with weaker edge clarity can use the maximum entropy threshold segmentation method, while the first type of images with higher edge clarity can use the iterative threshold method.

[0058] Preferably, after performing threshold segmentation on the image, an opening operation function may be called to smooth the contour of the measurement object, disconnect narrow necks, and eliminate thin protrusions.

[0059] As described above, threshold segmentation can divide an image into several non-overlapping regions according to grayscale features, and make these features present similarities in the same region, and present obvious differences between different regions. However, the threshold segmentation method only implements the segmentation of the image through a certain threshold value, without considering the spatial relationship of the image pixels, which may cause the segmented image to still contain some unnecessary interference information in addition to the measured object. According to an embodiment of the present invention, the image is secondary segmented by regional segmentation, specifically, the image is regionally segmented by regional growing method (S24), regardless of the type of the image.

[0060] Different from the threshold segmentation method, the region segmentation method takes the spatial properties of the image as the starting point, and directly finds pixels with certain similar properties in the image and divides them into sub-regions. The region growing method is a processing method that combines pixels or sub-regions into larger regions based on a certain pre-defined growth criterion. The basic principle of this method is: select certain specific pixels in the image as "seeds" and use them as the starting point for growth. Through a certain pre-defined similarity criterion, compare the pixels in the small area adjacent to the "seed" with the "seed". If they are similar enough, merge them into one category, and use the merged small area as a new "seed". Continue the above growth process until it stops growing, thus achieving the segmentation of the image.

[0061] That is, in the secondary segmentation, the measurement object is extracted from the image by the region growing method, the required measurement object is extracted according to the selected growth points, and batch analysis can be performed according to the number of selected growth points.

[0062] According to an embodiment of the present invention, after the image segmentation is completed, the shape information of the extracted measurement object is measured. In measuring the shape information of the extracted measurement object, a corresponding shape information measurement method is used according to the type of the image.

[0063] Specifically, after performing region segmentation on the image using the region growing method ( S24 ), it is determined whether the image is a third type image ( S31 ).

[0064] The third type of image (e.g. Figure 3A ) does not need to extract the edge of the measured object during measurement. However, the first type of image (e.g., Figure 1A ) and a second type of image (e.g., Figure 2A ) requires extracting the edge of the measured object during measurement. Figure 1A , Figure 2A and Figure 3A Different shape information measurement methods are explained in detail.

[0065] When it is determined that the image is the third type image (“Yes” in S31 ), the shape information of the extracted measurement object is directly measured.

[0066] by Figure 3A For example, all shape information of the object to be measured may include the diameter of the object. The diameter of the object indicates the diameter of the pimple. The diameter of the pimple can be used to evaluate the injection effect of a needle-free syringe.

[0067] Since the regionprops function can measure the properties of each connected region in the image (i.e., the adjacent regions with the same pixel value in the image), such as the area, center, and bounding box, the regionprops function (S32) can be directly called to measure the center coordinates and diameter of the measured object, i.e., the picoma diameter (S33).

[0068] In addition, the regionprops function can draw the bounding box of the measured object. Figure 3C shows the bounding box drawn with Figure 3A The superimposed image.

[0069] When it is determined that the image is not the third type of image ("No" in S31), that is, when it is determined that the image is the first type of image or the second type of image, the edge of the measurement object is extracted, a selection of a measurement start point and a measurement end point is received, and at least one of the measurement start point and the measurement end point is on the edge of the measurement object, and the distance between the measurement start point and the measurement end point is measured to measure at least a portion of the shape information of the measurement object.

[0070] by Figure 1A As an example, all shape information of the object to be measured includes: the length and width of the minimum circumscribed rectangle of the object, and the length from the upper end point of the object along the length direction of the object to a specific point inside the object (hereinafter referred to as the length of the puncture channel).

[0071] Figure 1A The final state of the injection of liquid medicine into a substance (e.g., gel) used to simulate skin by a needle-free syringe is shown. In the first stage of injection by the needle-free syringe, the liquid medicine is ejected and passes through the gel to form a cylindrical channel, which is called a deep hole. In the second stage, due to the elastic resistance of the gel, the incoming jet gathers near the hole depth, and the liquid medicine gathers to form a three-dimensional sphere. Then the liquid medicine gradually diffuses to the surroundings until the diffusion stops and the morphology no longer changes, so that the liquid medicine can be captured. Figure 1A .

[0072] Therefore, the length of the minimum circumscribed rectangle of the measurement object indicates the diffusion depth of the liquid medicine, and the width of the minimum circumscribed rectangle of the measurement object indicates the diffusion diameter of the liquid medicine. In addition to the diffusion depth and the diffusion diameter of the liquid medicine, since the depth of the hole is the actual puncture depth of the jet, the depth of the hole is also an indicator for evaluating the injection effect of the needle-free syringe, and the length of the puncture channel indicates the depth of the hole, and the end of the puncture channel (i.e., a specific point inside the measurement object) indicates the gathering point of the spherical solution. Among the shape information of these measurement objects, since the measurement start and end points of the puncture channel are inside the same connected area, the length of the puncture channel cannot be directly measured by calling the regionprops function.

[0073] Therefore, for Figure 1A The first type of image shown needs to extract the edge of the object to be measured. Specifically, the edge of the object to be measured is smoothed, and the edge of the object to be measured is extracted using the Canny edge extraction method (S33). Figure 1C Shows from Figure 1B The edge of the object to be measured is extracted. Figure 1D The extracted edge of the measured object is shown Figure 1A The superimposed image. Figure 1D As shown, there is a blue boundary line around the measurement object, and the blue boundary line is the edge of the measurement object.

[0074] When measuring the length of the puncture channel, as described above, the length of the puncture channel cannot be directly measured by the regionprops function, but needs to call the distance drawing function (S34) to receive the selection of the measurement start point and the measurement end point.

[0075] Specifically, a region of interest (ROI) can be set to measure the distance in the image, and the set ROI is in the form of a line. The line object is drawn by instantiating a function of the line object and combining it with a mouse click callback function, where the line object drawing function also sets two listeners to respond to line object clicks and movements. One listener is responsible for updating the label with the length of the line and changing the color of the line according to the length of the line, and the other listener is responsible for modifying and updating the known distance and unit of the measurement. In images with edges (for example, Figure 1D ) calls the function of drawing the line object. The user can select the starting point of the measurement by clicking the mouse, and the user can select the end point of the measurement by clicking the mouse again. The measured distance is in pixels by default. The user can double-click the line to pop up a dialog box so that the pixels can be changed to the required unit.

[0076] Preferably, the image may further include a reference object of known size, and the step of measuring the distance between the measurement start point and the measurement end point includes: receiving a modification of the measurement label scale, and calculating the actual distance between the measurement start point and the measurement end point according to the modified measurement label scale. Thus, the user can modify the measurement label scale according to the known size of the reference object, and when the reference object reaches the known size, the measurement label scale is set correctly at this time, so that the distance between the measurement start point and the measurement end point measured is the actual distance.

[0077] by Figure 1DAs an example, the measurement starting point of the puncture channel is on the edge of the measurement object (specifically, the upper end point), and the measurement end point of the puncture channel is inside the measurement object. The user can select the measurement end point of the puncture channel based on multiple judgment criteria such as the shrinkage trend from cylindrical to linear needle tip at the end of the channel, the starting point of the radial lines of gel cracking diverging to the surroundings, or the center point of the spherical solution bubble formed at the moment of injection. After receiving the selection of the measurement starting point and measurement end point of the puncture channel, the length of the puncture channel can be measured to be 30 mm, that is, the depth of the hole is 30 mm (S36).

[0078] When measuring the length and width of the minimum bounding rectangle of an object, you can directly call the regionprops function to extract the minimum bounding rectangle of the measured object and display the specific length and width in the image. For example, Figure 1D The bounding box of the minimum circumscribed rectangle of the measured object is shown, and the length and width of the minimum circumscribed rectangle of the measured object are 633 mm×882 mm.

[0079] by Figure 2A As an example, all shape information of the measurement object expected to be measured includes: the longest distance of the measurement object in the lateral direction and the longest distance in the longitudinal direction. The longest distance of the measurement object in the lateral direction indicates the intradermal injection diffusion width, and the longest distance of the measurement object in the longitudinal direction indicates the intradermal injection diffusion depth. The intradermal injection diffusion width and depth can be used to evaluate the injection effect of a needle-free syringe. The longest distance of the measurement object in the lateral direction and the longest distance in the longitudinal direction are not necessarily the length and width of the minimum circumscribed rectangle of the measurement object, and therefore cannot be measured by calling the regionprops function.

[0080] Figure 2A The treatment method and Figure 1A The processing method is the same as that of the embodiment, for example, extracting the edge of the measurement object (S34), Figure 2C Shows from Figure 2B The edge of the object to be measured is extracted. Figure 2D The extracted edge of the measured object is shown Figure 2A The superimposed image. Figure 2D As shown in FIG. 1 , there is a red boundary line around the measurement object, and the red boundary line is the edge of the measurement object. After extracting the edge of the measurement object ( S34 ), the distance drawing function is called to receive the selection of the measurement start point and the measurement end point ( S35 ). Figure 2DAs shown, the selected measurement starting point and measurement end point are both on the edge of the measurement object, and the measured distance between one set of measurement starting points and measurement end points is 3.4 mm, that is, the intradermal injection diffusion depth is 3.4 mm, and the distance between another set of measurement starting points and measurement end points is 10.0 mm, that is, the intradermal injection diffusion width is 10.0 mm (S36).

[0081] The image processing method for evaluating the injection effect of a needle-free syringe according to an embodiment of the present invention can automatically measure the shape information of the drug solution injected into a substance or an animal body by processing the image after the needle-free syringe test. The injection effect of the needle-free syringe can be evaluated based on the shape information, thereby achieving quantitative evaluation of the injection effect of the needle-free syringe.

[0082] The present invention also provides an image processing device for evaluating the injection effect of a needle-free syringe, which includes at least one data processor and at least one memory, and the at least one memory stores instructions. When the instructions are executed by the at least one data processor, the above-mentioned image processing method for evaluating the injection effect of a needle-free syringe is executed.

[0083] The various embodiments of the invention are not an exhaustive list of all possible combinations, but are intended to describe representative aspects of the invention, and what is described in various embodiments may be applied independently or in combinations of two or more.

[0084] The description presented in the above exemplary embodiments is only used to illustrate the technical solution of the present invention, and is not intended to be exhaustive, nor is it intended to limit the present invention to the precise form described. Obviously, it is possible for a person of ordinary skill in the art to make many changes and variations based on the above teachings. The exemplary embodiments are selected and described to explain the specific principles of the present invention and its practical application, so that other technicians in the field can easily understand, implement and use the various exemplary embodiments of the present invention and its various selected forms and modified forms. The scope of protection of the present invention is intended to be limited by the attached claims and their equivalent forms.

Claims

1. A method for processing an image for evaluating the injection effect of a needle-free syringe, wherein the image has a measurement object generated by a needle-free syringe test, and the method for processing an image for evaluating the injection effect of a needle-free syringe comprises the following steps: Preprocessing the image captured by the camera to remove irrelevant parts of the image and highlight key information in the image through image enhancement; Extracting the measurement object from the preprocessed image by combining threshold segmentation and region segmentation, and obtaining the measurement object by using a corresponding threshold segmentation method according to the type of the image; Measuring the extracted shape information of the measurement object, and using a corresponding shape information measurement method according to the type of the image to obtain at least a portion of the shape information of the measurement object; The images are classified into first type images, second type images and third type images, the first type images are images of injecting liquid medicine into a substance used to simulate skin, the second type images are cross-sectional images of the skin after injecting liquid medicine into the skin of an animal, and the third type images are images of skin papules generated after injecting liquid medicine into the skin of an animal; all shape information of the measuring object in the first type images includes: the length and width of the minimum circumscribed rectangle of the measuring object, the length of the upper end point of the measuring object along the length direction of the measuring object to a specific point inside the measuring object, the length of the minimum circumscribed rectangle of the measuring object indicates the diffusion depth of the liquid medicine, the minimum circumscribed rectangle of the measuring object The width indicates the diffusion diameter of the liquid medicine, and the length from the upper end point of the measuring object to a specific point inside the measuring object along the length direction of the measuring object indicates the depth of the hole; all the shape information of the measuring object in the second type of image includes: the longest distance in the transverse direction and the longest distance in the longitudinal direction of the measuring object, the longest distance in the transverse direction of the measuring object indicates the diffusion width of the intradermal injection, and the longest distance in the longitudinal direction of the measuring object indicates the diffusion depth of the intradermal injection; all the shape information of the measuring object in the third type of image includes the diameter of the measuring object, and the diameter of the measuring object indicates the diameter of the skin papule; and the steps of using the corresponding shape information measurement method according to the type of image include: determining whether the image is a third type of image; When it is determined that the image is a third type of image, directly measuring the extracted shape information of the measurement object; When it is determined that the image is not the third type of image, the edge of the measurement object is extracted, a selection of a measurement start point and a measurement end point is received, and at least one of the measurement start point and the measurement end point is on the edge of the measurement object, and the distance between the measurement start point and the measurement end point is measured to measure at least a portion of the shape information of the measurement object.

2. The image processing method for evaluating the injection effect of a needle-free syringe according to claim 1, wherein: The edge clarity of the measuring object in the first type of image is stronger than that in the second type of image and the third type of image. All shape information of the measuring object in the third type of image does not require extraction of the edge of the measuring object during measurement, while at least a portion of shape information of the measuring object in the first type of image and the second type of image requires extraction of the edge of the measuring object during measurement.

3. The image processing method for evaluating the injection effect of a needle-free syringe according to claim 2, wherein: The steps of using the corresponding threshold segmentation method according to the type of image include: determining whether the image is a first type of image; When it is determined that the image is a first type of image, performing threshold segmentation using an iterative threshold method; When it is determined that the image is not the first type of image, threshold segmentation is performed using a maximum entropy threshold segmentation method.

4. The image processing method for evaluating the injection effect of a needle-free syringe according to claim 3, wherein: After performing threshold segmentation on the image, the region growing method is used to perform region segmentation on the image.

5. The image processing method for evaluating the injection effect of a needle-free syringe according to claim 1, wherein: The step of directly measuring the extracted shape information of the measurement object includes: calling a regionprops function to measure the shape information of the measurement object.

6. The image processing method for evaluating the injection effect of a needle-free syringe according to claim 1, wherein: The step of extracting the edge of the measured object includes: smoothing the edge of the measured object and extracting the edge of the measured object using a canny edge extraction method.

7. The image processing method for evaluating the injection effect of a needle-free syringe according to claim 1, wherein: The step of receiving selections of a measurement start point and a measurement end point includes: calling a distance drawing function, receiving selections of a measurement start point through a mouse click by the user, and receiving selections of a measurement end point through a mouse click by the user again.

8. The image processing method for evaluating the injection effect of a needle-free syringe according to claim 1, wherein: The image further includes a reference object of known size, The step of measuring the distance between the measurement start point and the measurement end point includes: receiving a modification of the measurement tag scale, and calculating the actual distance between the measurement start point and the measurement end point according to the modified measurement tag scale.

9. The image processing method for evaluating the injection effect of a needle-free syringe according to claim 1, wherein: The steps of image preprocessing include: receiving an image captured by a camera and displaying the image; Crop the image; Perform RGB separation on the cropped image; The image is enhanced by adjusting the contrast and / or grayscale value of the separated G channel image or R channel image.

10. The image processing method for evaluating the injection effect of a needle-free syringe according to claim 9, wherein: The steps to perform RGB separation on the cropped image include: Determine whether the liquid injected during the needleless syringe test is rhodamine B dye; When it is determined that the liquid injected during the needle-free syringe test is a rhodamine B dye, the cropped image is separated into a G channel image; When it is determined that the drug solution injected in the needle-free syringe test is not the rhodamine B dye, the cropped image is separated into an R channel image.

11. A device for processing images for evaluating injection effects of a needle-free injector, comprising: at least one data processor; as well as At least one memory storing instructions, which, when executed by the at least one data processor, enable execution of the image processing method for evaluating the injection effect of a needle-free injector according to any one of claims 1 to 10.

Images