Intelligent processing method for burn skin tissue dyeing result
By adopting multi-threshold edge detection and region growth algorithms in burned skin tissue staining, the growth threshold is adaptively adjusted, which solves the problem of unclear cell morphological characteristics after staining, and achieves higher precision cell segmentation.
Patent Information
- Application Number
- CN202510912558.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-03
- Publication Date
- 2025-08-01
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
During the operation, the existing burn skin tissue staining methods are unclear in the morphological characteristics of cells after staining due to hypoxia in the material, extrusion and dye diffusion, which affects the normal analysis of the state of nerve fibroblasts.
An intelligent processing method is adopted to perform edge detection and morphological operations using thresholds of different sizes, obtain edge images of the stained image, filter the target coordinate position, adjust the degree of edge chaos, combine the region growth algorithm to segment the stained image, and adaptively adjust the growth threshold to improve the segmentation accuracy.
It improves the segmentation accuracy and robustness of burn skin tissue staining results, enhances the capture of cell nucleus and cell membrane edge information, and achieves more accurate cell segmentation.
Smart Images
Figure CN120411104A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of image data processing, and particularly relates to an intelligent processing method for the staining results of burn skin tissue. Background Art
[0002] Related research shows that cutaneous sensory nerve fibers participate in wound healing by secreting neuropeptides, and the number and morphological states of nerve fibers in granulation tissue at different stages of the wounds of burn patients are different. The internal structures and pathological states of nerve fibers in the granulation tissue of the wound are different within 1, 2, 3, and 4 weeks after burn.
[0003] In order to more clearly observe the two-dimensional and three-dimensional spatial structures of nerve fibers and understand the morphological changes of nerve fibers, microscopic examination of the wounds of burn patients is required. At this time, fluorescent chemical substances are used to stain nerve fiber cells, but the cell staining operation is relatively complex. During the staining operation, due to situations such as lack of oxygen in the sample collection, extrusion, and diffusion of the staining agent, the morphological characteristics of the cells after staining are not clear, which is not conducive to the normal analysis of the state of nerve fiber cells in the wounds of burn patients. Summary of the Invention
[0004] The present invention provides an intelligent processing method for the staining results of burn skin tissue to solve the existing problems.
[0005] The intelligent processing method for the staining results of burn skin tissue of the present invention adopts the following technical solutions: An embodiment of the present invention provides an intelligent processing method for the staining results of burn skin tissue, and the method includes the following steps: Obtain the stained image of the burn skin tissue; Use a number of thresholds of different sizes to perform edge detection and morphological operations on the stained image to obtain edge images under each threshold, obtain the gradient direction of each pixel point in the stained image, obtain the first edge possibility of the pixel point according to the gradient direction of each pixel point and the gray values of all edge pixel points within a preset local range, screen out the target coordinate positions among the coordinate positions of all pixel points in the edge image under each threshold according to the differences between the edge images under different thresholds, and obtain the edge chaos degree at any target coordinate position of the edge image under each threshold according to the length of the edge line, the number of edge pixel points, and the gray value in different edge images; use the number of edge lines in the edge image to adjust the edge chaos degree at the target coordinate position in the edge images under all thresholds to obtain the multi-threshold chaos degree of the target coordinate position; use the multi-threshold chaos degree to adjust the first edge possibility of the pixel points in the stained image to obtain the second edge possibility of the pixel points in the stained image; The initial growth threshold of the preset region growth algorithm is adjusted using the second edge possibility of each pixel point in the stained image to obtain the new growth threshold of each pixel point; Perform region growth processing on the stained image in combination with the new growth threshold to obtain the segmentation result image of the stained image.
[0006] Further, the method of using several thresholds of different sizes to perform edge detection and morphological operations on the stained image to obtain the edge image under each threshold specifically includes: Preset a threshold interval , preset As the iteration step size and obtain several thresholds in the threshold interval from small to large. Denote the sequence formed by all the obtained thresholds as the threshold sequence. Combine the Sobel operator to perform edge detection to obtain the initial edge image under each threshold and perform thinning processing of morphological operations on the initial edge image under each threshold to obtain the edge image under each threshold.
[0007] Further, the method of obtaining the gradient direction of each pixel point in the stained image and obtaining the first edge possibility of the pixel point according to the gradient direction of each pixel point and the gray values of all edge pixel points in a preset local range specifically includes: Use the Sobel operator to obtain the gradient direction of each pixel point in the stained image; In the edge image under any threshold, construct a window centered on any pixel point. Preset the size of the window as , and denote the variance of the gray values corresponding to all edge pixel points in the window in the stained image as the gray chaos degree of the pixel point at the center of the window; Denote the product result of the gray chaos degree of the th pixel point in the stained image and the cosine value of the gradient direction of the th pixel point as the first edge possibility of the th pixel point in the stained image under the corresponding threshold.
[0008] Further, the method of screening out the target coordinate positions among the coordinate positions of all pixel points in the edge image under each threshold according to the differences between the edge images under thresholds of different sizes specifically includes: Obtain the absolute value of the difference between the gray values of the pixel points at the same coordinate position in the edge images corresponding to the rd threshold and the th threshold, and denote it as the edge change parameter of the coordinate position. Denote the coordinate position with a non-zero edge change parameter as the target coordinate position of the edge image under the th threshold.
[0009] Further, obtaining the edge chaos degree at any target coordinate position of the edge image at each threshold according to the lengths of the edge lines, the number of edge pixels, and the gray values in different edge images includes the following specific methods: At the th threshold and the th threshold, respectively corresponding edge images, record the length of the edge line to which any target coordinate position belongs as the relative edge length of the target coordinate position; record the edge lines in the edge image at the th threshold that have no intersection in the coordinate position with all edge lines in the edge image at the th threshold as the newly added edge lines in the edge image at the th threshold, and obtain several newly added edge lines in the edge image at the th threshold; record the average gray value of the edge pixels on all edge lines in the edge image at the th threshold as the original edge gray parameter of the th threshold. The specific calculation method for the edge chaos degree at any target coordinate position of the edge image at the th threshold is as follows: Among them, represents the edge chaos degree of the target coordinate position of the edge image at the th threshold; represents the relative edge length of the target coordinate position; represents the number of edge pixels in the edge image at the th threshold; represents the average gray value of all edge pixels on the th newly added edge line in the edge image at the th threshold; represents the number of newly added edge lines in the edge image at the th threshold; represents the original edge gray parameter of the th threshold; represents the chaos parameter of the th threshold; represents obtaining the absolute value.
[0010] Further, the specific calculation method for the chaos parameter of the th threshold is as follows: .
[0011] Further, the method for adjusting the edge disorder degree at the target coordinate position in the edge images under all thresholds by using the number of edge lines in the edge image to obtain the multi-threshold disorder degree at the target coordinate position includes the following specific steps: For the edge image under the -th threshold relative to the edge image under the -th threshold, the edge lines whose edge lines have intersections in the coordinate positions and the coordinate positions of all edge pixels in the edge lines are not completely the same are recorded as the changing edge lines in the edge image under the -th threshold; the cumulative value of the number of changing edge lines in the edge images under all thresholds is recorded as the first cumulative value, and the ratio of the number of changing edge lines in the edge image under the -th threshold to the first cumulative value is recorded as the first parameter of the -th threshold, and linear normalization is performed on the first parameters of all thresholds to obtain the weight factor of each threshold; The specific calculation method for the multi-threshold disorder degree at the target coordinate position is as follows: where, represents the multi-threshold disorder degree at the target coordinate position; represents the weight factor under the -th threshold; represents the edge disorder degree at the target coordinate position in the edge image under the -th threshold; represents the number of all thresholds in the threshold sequence.
[0012] Further, the method for adjusting the first edge possibility of the pixel points in the stained image by using the multi-threshold disorder degree to obtain the second edge possibility of the pixel points in the stained image includes the following specific steps: The cumulative value of the first edge possibility of the pixel point in the edge images under all thresholds is recorded as the first value, and the product of the multi-threshold disorder degree when the coordinate position of the pixel point in the stained image is the target coordinate position in the edge image and the first value of the pixel point is recorded as the second edge possibility of the pixel point.
[0013] Further, the method for adjusting the initial growth threshold by using the second edge possibility of each pixel point in the stained image to obtain the new growth threshold of each pixel point includes the following specific steps: The specific calculation method for the new growth threshold of any pixel point in the stained image is as follows: where, represents the new growth threshold; represents the initial growth threshold; Represents the second edge possibility of a pixel point; Represents a linear normalization function.
[0014] Furthermore, performing region growing processing on the stained image by combining the new growth threshold to obtain a segmentation result image of the stained image, the specific method includes: Performing segmentation processing on the stained image according to the absolute value of the difference in gray values of different pixel points and using the region growing algorithm. During the region growing process, when the absolute value of the difference in gray values of a pixel point in the stained image exceeds the new growth threshold, stop region growing at the corresponding position of the pixel point, and record the image obtained after region growing is completed as the segmentation result image of the stained image.
[0015] The beneficial effects of the technical solution of the present invention are as follows: Adjusting the initial growth threshold using the second edge possibility of each pixel point in the stained image, correcting the initial growth threshold according to the second edge possibility of the pixel point to obtain the new growth threshold of each pixel point. This adaptive adjustment method can flexibly adjust the growth threshold according to the actual situation, making the region growing algorithm more adaptable to different nuclear shapes, sizes, and edge features, thereby improving the accuracy of the segmentation result. In addition, through edge detection under multiple thresholds, obtaining the edge confusion degree between the nucleus and the cell membrane, and determining the growth threshold at different positions according to the edge detection result, thereby realizing the segmentation of immunostained tissue cells, effectively enhancing the capture of nuclear edge information, contributing to more accurate segmentation of the nucleus and the cell membrane, and improving the segmentation accuracy and robustness. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.
[0017] Figure 1 It is a step flow chart of an intelligent processing method for the staining result of burn skin tissue of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0018] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following will, in combination with the drawings and preferred embodiments, describe in detail the specific implementation manner, structure, features, and effects of an intelligent processing method for the staining result of burn skin tissue proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.
[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs.
[0020] The following specifically describes the specific solution of an intelligent processing method for the staining results of burn skin tissue provided by the present invention in conjunction with the accompanying drawings.
[0021] Please refer to Figure 1 , which shows a flowchart of the steps of an intelligent processing method for the staining results of burn skin tissue provided by an embodiment of the present invention. The method includes the following steps: Step S001: Obtain a stained image of the burn skin tissue.
[0022] It should be noted that during the chemical staining process of the burn skin tissue, due to the diffusion of the dye reagent, the boundary between the dye reagent and the tissue cells in the burn skin tissue cannot be clearly defined, resulting in the inability to effectively and accurately segment the stained cells, affecting the subsequent detection results. Therefore, in order to improve the accurate segmentation of the cells after chemical staining of the burn skin tissue, it is first necessary to obtain the corresponding image after staining the burn skin tissue.
[0023] Specifically, in order to implement an intelligent processing method for the staining results of burn skin tissue proposed in this embodiment, it is first necessary to collect. The specific process is as follows: Use a digital microscope to obtain an image of the stained burn skin tissue, denoted as a stained image.
[0024] So far, the stained image of the burn skin tissue is obtained through the above method.
[0025] Step S002: Edge detection and morphological operations are performed on the stained image using a number of thresholds of different sizes to obtain edge images under each threshold. The gradient direction of each pixel point in the stained image is obtained. According to the gradient direction of each pixel point and the gray values of all edge pixel points within a preset local range, the first edge possibility of the pixel point is obtained. According to the differences between the edge images under thresholds of different sizes, the target coordinate positions among the coordinate positions of all pixel points in the edge image under each threshold are screened out. According to the length of the edge line, the number of edge pixel points, and the gray value in the edge images of different thresholds, the edge disorder degree at any target coordinate position in the edge image under each threshold is obtained. According to the position conditions of the edge lines in the edge images under different thresholds, the changing edge lines in the edge image under each threshold are obtained. The edge disorder degree at the target coordinate position in the edge images under all thresholds is adjusted using the number of edge lines in the edge images, and the multi-threshold disorder degree of the target coordinate position is obtained. The first edge possibility of the pixel points in the stained image is adjusted using the multi-threshold disorder degree to obtain the second edge possibility of the pixel points in the stained image.
[0026] It should be noted that for the images of stained burn skin tissue cells, the cell nucleus, as the stained area, has a high contrast and obvious edges, so it often presents relatively prominent features in the image. The cell membrane edge of the cell where the stained cell nucleus is located is clearly shown, but when performing cell edge detection, the edge of the cell nucleus will affect the edge of the cell membrane.
[0027] Generally, the smoothness of the cell nucleus edge is low, resulting in a high degree of disorder in its edge detection results, and the edge changes detected under different thresholds are also relatively chaotic. Therefore, in this embodiment, edge detection is performed under multiple thresholds to obtain the edge disorder degrees of the cell nucleus and the cell membrane, and then according to the edge detection results, the growth thresholds at different positions are determined, thereby realizing the segmentation of immunostained tissue cells.
[0028] Specifically, in step (2.1), first, a preset threshold interval , and a preset is used as the iteration step size, and several thresholds in the threshold interval are obtained from small to large. The sequence formed by all the obtained thresholds is denoted as the threshold sequence. Edge detection is combined with the Sobel operator to obtain the initial edge image under each threshold, and a thinning process of morphological operations is performed on the initial edge image under each threshold to obtain the edge image under each threshold.
[0029] It should be noted that in the preset threshold interval according to experience , , , which can be adjusted separately according to the actual situation, and is not specifically limited in this embodiment.
[0030] It should be noted that by refining the edge image in this embodiment, the edge lines in the edge image can be made more detailed and continuous, facilitating the subsequent more accurate acquisition of the actual edge lines in the stained image, so as to further accurately segment the cells in the stained image of the burned skin tissue.
[0031] It should be noted that since the Sobel operator is an existing edge detection operator, it will not be specifically described in this embodiment.
[0032] It should be noted that the edge confusion degree refers to the complexity and confusion degree of the edges in the image. According to the edge detection information under multiple thresholds, the edge confusion degree of the cell edges at the same position can be obtained, and the growth conditions at different positions can be adjusted according to the size of the cell edge confusion degree, and then the pixel points with similar properties can be merged and grown to better control the growth range and results.
[0033] It should be noted that since the edge detection results at the same position are different in all the edge images obtained by edge detection under different thresholds, in order to accurately obtain the edge information of the cell nucleus in the stained image, this embodiment analyzes the gray distribution characteristics of the pixel points in the neighborhood range to obtain the possibility of the pixel points in the stained image being the cell nucleus edge.
[0034] Then, obtain the first edge possibility of the pixel points in the stained image. As an embodiment, the specific calculation method is as follows: In the edge image at any threshold, construct a window centered on any pixel point in the stained image, and the preset size of the window is , denote the variance of the gray values corresponding to all the edge pixel points in the window in the stained image as the gray confusion degree of the pixel point at the center of the window, and obtain the first edge possibility of each pixel point in the stained image in the edge image at the corresponding threshold according to the gray confusion degree and gradient direction of the pixel point. As an embodiment, the specific calculation method is as follows: Among them, represents the first edge possibility of the th pixel point in the stained image at the corresponding threshold, represents the gray confusion degree of the th pixel point; represents the cosine value of the gradient direction of the th pixel point.
[0035] It should be noted that according to experience, the preset parameter is 9, which can be adjusted according to the actual situation and is not specifically limited in this embodiment.
[0036] It should be noted that the first edge possibility of a pixel reflects the magnitude of the possibility of the pixel in the stained image being the edge of the cell nucleus. The greater the first edge possibility, the greater the possibility of the pixel in the stained image being the edge of the cell nucleus.
[0037] It should be noted that at different thresholds of edge detection, the degree of edge chaos corresponding to the same position is different. At the same position, the more chaotic the gray values of the pixels within the window, the more chaotic the edge it reflects, and the greater the possibility that the pixel is located in the cell nucleus; the greater the difference in the connection directions between any adjacent pixels, that is, the worse the continuity of the edge and the more inconsistent the formed edge directions, the smaller the possibility that it belongs to the edge of the cell nucleus.
[0038] It should be noted that since the larger the corresponding threshold during edge detection, the fewer the number of pixels at the same position in the edge images at all thresholds, the more obvious the change in the number of edge pixels at the same corresponding position in the edge images obtained under different thresholds, that is, the more chaotic the change in the edge information in the stained image. When the threshold during edge detection increases or decreases, the edge lines in the edge image generally change in the form of short line segments. When the threshold of edge detection changes, the more obvious the change in the corresponding edge lines at the same position before and after the change indicates that the gray values of the edge pixels in the edge lines at the corresponding threshold are closer, and the degree of chaos of the pixels at the corresponding position is smaller, otherwise, a greater degree of edge chaos will be shown. Therefore, the degree of edge change chaos is reflected by the edge lengths at the same corresponding position under different thresholds.
[0039] Step (2.2), obtain the absolute value of the difference in gray values between the pixels at the same coordinate position in the edge images corresponding to the th threshold and the th threshold, and denote it as the edge change parameter of the coordinate position. Denote the coordinate positions where the edge change parameter is not 0 as the target coordinate positions of the edge image at the th threshold; in the edge images corresponding to the th threshold and the th threshold, denote the length of the edge line to which any target coordinate position belongs as the relative edge length of the target coordinate position; denote the edge lines in the edge image at the th threshold that have no intersection in the coordinate position with all the edge lines in the edge image at the th threshold as the newly added edge lines in the edge image at the th threshold, and obtain several newly added edge lines in the edge image at the th threshold; denote the average gray value of the edge pixels on all the edge lines in the edge image at the th threshold as the The original edge grayscale parameter of a threshold.
[0040] Step (2.3), first, obtain The degree of edge chaos at any target coordinate position of the lower-edge image at the where, represents the degree of edge chaos at the target coordinate position of the lower-edge image at the th threshold; represents the relative edge length of the target coordinate position; represents the th number of edge pixel points in the lower-edge image at the th threshold; represents the th average grayscale value of all edge pixel points on the newly added edge line in the lower-edge image at the th threshold; represents the number of newly added edge lines in the lower-edge image at the th threshold; represents the original edge grayscale parameter of the th threshold; [[ID=C37]]
[0041] It should be noted that since there is no edge image for comparison in the edge image at the minimum threshold in the threshold sequence, the degree of edge chaos at any coordinate position at the minimum threshold is set to 0 in this embodiment. The specific situation can be adjusted according to actual needs, and no specific limitation is made in this embodiment.
[0042] It should be noted that the original edge grayscale parameter of each threshold reflects the overall grayscale level of all edge pixel points in the edge image at the previous threshold of this threshold. By obtaining the original edge grayscale parameter and the th average grayscale value of all edge pixel points on the newly added edge line at the The difference reflects the difference in the gray value levels of the upper-edge pixel points on the newly added edge line after the threshold change relative to the gray value levels of the upper-edge pixel points on the edge line before the threshold change. The greater the difference, the greater the difference in the gray value of the newly added edge line generated after the threshold change in the stained image compared to the gray value corresponding to the previous threshold, indicating that the gray distribution of the newly added edge line is more chaotic overall; the relative edge length reflects the length of the edge line corresponding to the edge line whose length changes at the same position in the edge images under adjacent thresholds of different sizes in the threshold sequence. The smaller the length, the less able the edge line is to completely correspond to the edge of the cell nucleus in the stained image, that is, the gray values of all the edge pixel points included when the edge line is complete are more inconsistent. Therefore, the degree of edge chaos at the target coordinate position belonging to this edge line is greater.
[0043] Then, for the edge image under the -th threshold relative to the edge image under the -th threshold, the edge lines whose edge lines have intersections in the coordinate positions and the coordinate positions of all the edge pixel points in the edge lines are not completely the same are recorded as the changing edge lines in the edge image under the -th threshold; the degree of edge chaos at any target coordinate position in the edge images under different thresholds is weighted and adjusted to obtain the multi-threshold chaos degree of the target coordinate position. As an example, the specific calculation method is: where, represents the multi-threshold chaos degree of the target coordinate position; represents the weight factor under the -th threshold; represents the number of changing edge lines in the edge image under the -th threshold; represents the degree of edge chaos at the target coordinate position of the edge image under the -th threshold; represents the number of all thresholds in the threshold sequence; represents the linear normalization function.
[0044] It should be noted that since there is no -th threshold when the minimum threshold is the -th threshold in the threshold sequence, there is no newly added edge line in the edge image under the minimum threshold. Therefore, when obtaining the weight factor of the threshold, the cumulative value of the number of changing edge lines in the edge images under the 2nd threshold to the V-th threshold in the threshold sequence is obtained, that is , to represent the total number of edge lines whose lengths change during the process of all thresholds changing before and after.
[0045] Finally, obtain the second edge possibility of any pixel point in the stained image. As an example, the specific calculation method is as follows: Wherein, represents the second edge possibility of the pixel point in the stained image; represents the multi-threshold confusion degree when the coordinate position of the pixel point in the stained image is the target coordinate position in the edge image; represents the first edge possibility of the pixel point in the edge image under the th threshold; represents the number of all thresholds in the threshold sequence.
[0046] It should be noted that the second edge possibility of the pixel point corresponding to the target coordinate position in the stained image is used to describe the probability that the pixel point is the nucleus edge in the stained image. The greater the second edge possibility, the greater the probability that the corresponding pixel point is the nucleus edge.
[0047] It should be noted that since the target coordinate position represents the coordinate position corresponding to the edge pixel points where the edge information changes under several different sizes of thresholds, the pixel points at the coordinate positions different from the target coordinate position in the stained image cannot be edge pixel points, and there is no multi-threshold confusion degree at the coordinate positions different from the target coordinate position in the stained image, that is, the multi-threshold confusion degree at this coordinate position is 0. Therefore, in this embodiment, the second edge possibility of the pixel points in the stained image whose coordinate positions do not correspond to the target coordinate position is also set to 0.
[0048] So far, the second edge possibility of each pixel point in the stained image is obtained through the above method.
[0049] Step S003: Preset the initial growth threshold of the region growing algorithm, and adjust the initial growth threshold by using the second edge possibility of each pixel point in the stained image to obtain the new growth threshold of each pixel point.
[0050] Specifically, for the initial growth threshold A of the preset region growing algorithm, the initial growth threshold is corrected according to the second edge possibility of the pixel point to obtain the new growth threshold of any pixel point in the stained image. As an example, the specific calculation method is as follows: Wherein, represents the new growth threshold; represents the initial growth threshold; represents the second edge possibility of the pixel point; represents the linear normalization function.
[0051] It should be noted that according to experience, the initial growth threshold of the region growing algorithm is 20, which can be adjusted according to specific situations and is not specifically limited in this embodiment.
[0052] It should be noted that by performing edge detection processing under multiple thresholds, the edge confusion degree is obtained, and the growth threshold at different positions is determined accordingly. Then, points with similar properties are merged and continue to grow outward until no more pixels that meet the conditions are included. This operation improves the determination of the edge pixels of the stained cells and the region to be segmented, and improves the segmentation accuracy.
[0053] Thus far, the new growth threshold of any pixel point in the stained image is obtained through the above method.
[0054] Step S004: Perform region growing processing on the stained image in combination with the new growth threshold to obtain the segmentation result image of the stained image.
[0055] Specifically, the stained image is segmented by using the region growing algorithm according to the absolute value of the difference in the gray values of different pixel points. During the region growing process, when the absolute value of the difference in the gray values of the pixel points in the stained image exceeds the new growth threshold, the region growing stops at the corresponding position of the pixel point, and the image obtained after the region growing is completed is recorded as the segmentation result image of the stained image.
[0056] Thus far, this embodiment is completed.
[0057] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention shall be included within the protection scope of the present invention.
Claims
1. An intelligent processing method for the staining results of burned skin tissue, characterized in that, The method includes the following steps: Obtain a stained image of the burned skin tissue; Perform edge detection and morphological operations on the stained image using a number of thresholds of different sizes to obtain edge images under each threshold, obtain the gradient direction of each pixel point in the stained image, and obtain the first edge possibility of the pixel point according to the gradient direction of each pixel point and the gray values of all edge pixel points within a preset local range. According to the differences between the edge images under thresholds of different sizes, screen out the target coordinate positions among the coordinate positions of all pixel points in the edge image under each threshold. According to the lengths of the edge lines, the number of edge pixel points, and the gray values in different edge images, obtain the edge confusion degree at any target coordinate position of the edge image under each threshold; adjust the edge confusion degree at the target coordinate position in the edge images under all thresholds using the number of edge lines in the edge image to obtain the multi-threshold confusion degree of the target coordinate position; adjust the first edge possibility of the pixel points in the stained image using the multi-threshold confusion degree to obtain the second edge possibility of the pixel points in the stained image; Preset the initial growth threshold of the region growing algorithm, and adjust the initial growth threshold using the second edge possibility of each pixel point in the stained image to obtain the new growth threshold of each pixel point; Perform region growing processing on the stained image in combination with the new growth threshold to obtain the segmentation result image of the stained image.
2. The intelligent processing method for the staining result of burned skin tissue according to claim 1, wherein, The specific method included in performing edge detection and morphological operations on the stained image using a number of thresholds of different sizes to obtain edge images under each threshold is as follows: Preset threshold range , preset as the iteration step size and obtain several thresholds in the threshold range from small to large. Denote the sequence formed by all the obtained thresholds as the threshold sequence. Combine with the Sobel operator to perform edge detection to obtain the initial edge image under each threshold and perform thinning processing of morphological operations on the initial edge image under each threshold to obtain the edge image under each threshold.
3. The intelligent processing method for the staining result of burned skin tissue according to claim 1, wherein, The specific method included in obtaining the gradient direction of each pixel point in the stained image and obtaining the first edge possibility of the pixel point according to the gradient direction of each pixel point and the gray values of all edge pixel points within a preset local range is as follows: Use the Sobel operator to obtain the gradient direction of each pixel point in the stained image; In the edge image at any threshold, construct a window centered on any pixel point, and preset the size of the window to be , and denote the variance of the gray values corresponding to all edge pixel points in the window in the staining image as the gray chaos degree of the pixel point at the center of the window; denote the product result of the gray chaos degree of the -th pixel point in the staining image and the cosine value of the gradient direction of the -th pixel point as the first edge possibility of the -th pixel point in the staining image at the corresponding threshold.
4. The intelligent processing method for the staining result of burned skin tissue according to claim 1, wherein, The specific method included in screening out the target coordinate positions among the coordinate positions of all pixel points in the edge image under each threshold according to the differences between the edge images under thresholds of different sizes is as follows: Obtain the th threshold and the th threshold, respectively, and calculate the absolute value of the difference in grayscale values between the pixel points at the same coordinate positions in the edge images corresponding to them. Denote this as the edge change parameter at the coordinate position. Denote the coordinate positions where the edge change parameter is not 0 as the target coordinate positions of the edge image under the th threshold.
5. The intelligent processing method for the staining result of burn skin tissue according to claim 1, wherein The specific method included in obtaining the edge confusion degree at any target coordinate position of the edge image under each threshold according to the lengths of the edge lines, the number of edge pixel points, and the gray values in different edge images is as follows: Among the edge images corresponding to the nd threshold and the th threshold respectively, the length of the edge line to which any target coordinate position belongs is denoted as the relative edge length of the target coordinate position; the edge lines in the edge image under the th threshold that have no intersection in the coordinate position with all the edge lines in the edge image under the th threshold are denoted as the newly added edge lines in the edge image under the th threshold, and several newly added edge lines in the edge image under the th threshold are obtained; the average gray value of the edge pixel points on all the edge lines in the edge image under the th threshold is denoted as the original edge gray parameter of the th threshold; The specific calculation method for the degree of edge chaos at any target coordinate position of the lower edge image of the threshold is as follows: Among them, represents the edge disorder degree of the target coordinate position of the edge image under the th threshold; represents the relative edge length of the target coordinate position; represents the number of edge pixel points in the edge image under the th threshold; represents the average gray value of all edge pixel points on the th new edge line in the edge image under the th threshold; represents the number of new edge lines in the edge image under the th threshold; represents the original edge gray parameter of the th threshold; represents the disorder parameter of the th threshold; represents obtaining the absolute value.
6. The intelligent processing method for the staining result of burn skin tissue according to claim 5, wherein The specific calculation method of the chaos parameter of the th threshold is as follows: .
7. The intelligent processing method for the staining result of burned skin tissue according to claim 1, wherein The specific method included in adjusting the edge confusion degree at the target coordinate position in the edge images under all thresholds using the number of edge lines in the edge image to obtain the multi-threshold confusion degree of the target coordinate position is as follows: Among the edge images under the th threshold and the edge images under the th threshold, the edge lines whose edge lines have intersections in the coordinate positions and the coordinate positions of all edge pixels in the edge lines are not exactly the same are recorded as the changing edge lines in the edge images under the th threshold; the cumulative value of the number of changing edge lines in the edge images under all thresholds is recorded as the first cumulative value, and the ratio of the number of changing edge lines in the edge image under the th threshold to the first cumulative value is recorded as the first parameter of the th threshold, and linear normalization is performed on the first parameters of all thresholds to obtain the weight factor of each threshold; The specific calculation method of the multi-threshold confusion degree of the target coordinate position is: Among them, represents the multi-threshold chaos degree of the target coordinate position; represents the weight factor under the th threshold; represents the edge chaos degree of the target coordinate position of the edge image under the th threshold; represents the number of all thresholds in the threshold sequence.
8. The intelligent processing method for the staining result of burned skin tissue according to claim 1, characterized in that, The specific method included in adjusting the first edge possibility of the pixel points in the stained image using the multi-threshold confusion degree to obtain the second edge possibility of the pixel points in the stained image is as follows: The cumulative value of the first edge possibility of a pixel in the edge images at all thresholds is denoted as the first value. The product of the multi-threshold confusion degree when the coordinate position of the pixel in the staining image is the target coordinate position in the edge image and the first value of the pixel is denoted as the second edge possibility of the pixel.
9. The intelligent processing method for the staining result of burn skin tissue according to claim 1, characterized in that The method of adjusting the initial growth threshold by using the second edge possibility of each pixel in the staining image to obtain the new growth threshold for each pixel specifically includes: The specific calculation method of the new growth threshold for any pixel in the staining image is: Among them, represents the new growth threshold; represents the initial growth threshold; represents the second edge possibility of the pixel point; represents the linear normalization function.
10. The intelligent processing method for the staining result of burned skin tissue according to claim 1, wherein, The method of performing region growing processing on the staining image by combining the new growth threshold to obtain the segmentation result image of the staining image specifically includes: Performing segmentation processing on the staining image according to the absolute value of the difference in gray values of different pixels and using the region growing algorithm. During the region growing process, when the absolute value of the difference in gray values of a pixel in the staining image exceeds the new growth threshold, region growing stops at the corresponding position of the pixel, and the image obtained after the region growing is completed is denoted as the segmentation result image of the staining image.