Three-dimensional reconstruction method of microstructure of pineapple leaf fiber

Abstract

The application provides a three-dimensional reconstruction method of a pineapple leaf fiber microstructure, relates to the technical field of three-dimensional reconstruction of fiber microstructure, and comprises the following steps: an equivalent circle method is used to calculate the equivalent diameter of a single fiber branch and the single pore diameter, area normalization is used to calculate the area density of the corresponding pore area, a screening condition is set to screen the area density, and an effective pore area is obtained; coordinate transformation is used to obtain the center point coordinates of each pore in the effective pore area; for the center point coordinates of each pore, the radial distance of the center point coordinates to the fiber center line is calculated, and the center point coordinates are excluded; a cylindrical coordinate system is established, and an equivalent cylindrical model is used to reconstruct the three-dimensional volume of the area occupied by each pore. The area normalization density calculation and the dynamic setting of the screening condition according to the global statistical parameter improve the physical rationality of the pore distribution; and the cylindrical coordinate system is established based on the fiber center line, and is more suitable for the geometric characteristics of the natural fiber.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of three-dimensional reconstruction of fiber microstructure, in particular to a three-dimensional reconstruction method of pineapple leaf fiber microstructure. BACKGROUND

[0002] Traditional fiber structure analysis methods have significant limitations in analyzing complex porous materials such as pineapple leaf fibers. They mostly rely on two-dimensional image analysis, which is difficult to accurately distinguish the topological relationship between fiber branches and pore regions, resulting in large statistical deviations in pore size distribution. Especially in heterogeneous fiber cross-sections, conventional threshold segmentation is easily disturbed by noise, and it is difficult to accurately identify fiber centerlines and associated pore diameters. Three-dimensional structure reconstruction often relies on assumed models, ignoring real spatial distribution.

[0003] In the prior art, CN112991518A discloses a method for extracting the coordinates of the central axis of the fiber in three-dimensional space, calculating the radius of the fiber, and thereby reconstructing the three-dimensional model of the fiber. The method uses a boundary tracking algorithm to mark connected domains and connects the connected domains to repair the missing parts of the fiber caused by occlusion. However, this method is only suitable for fiber reconstruction of textiles with regular distribution, and does not consider the decisive role of micro-porous structure in plant respiration, the pore size distribution density of plant fibers, the geometric properties of fiber branches and pores, and the non-uniform expansion characteristics of pores in the fiber cross-section. Therefore, there is an urgent need for a three-dimensional reconstruction method for pineapple leaf fiber microstructure.

[0004] The above information disclosed in the background section is only intended to enhance the understanding of the background of the present disclosure, and therefore it can include information that does not constitute the prior art known to those of ordinary skill in the art. SUMMARY

[0005] The present application aims to provide a three-dimensional reconstruction method for pineapple leaf fiber microstructure to solve the problems raised in the background.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0007] The three-dimensional reconstruction method for pineapple leaf fiber microstructure includes the following specific steps:

[0008] S1: Collect the scanning electron microscope image of the pineapple leaf cross-section and label it as the first image. Based on the pre-set segmentation threshold, perform background segmentation on the first image to identify the fiber centerline, each fiber branch region, and the corresponding pore region, and label the identified image as the second image.

[0009] S2: Based on the second image, the number of pixels of each fiber branch region and the internal single aperture is counted respectively, and the equivalent circle method is used to calculate the equivalent diameter of the single fiber branch and the single aperture diameter respectively, based on the single aperture diameter, the area density of the corresponding aperture region is calculated by area normalization, and the expectation and standard deviation of the aperture region corresponding to each fiber branch region of all second images are calculated, the screening condition is set to screen the area density, and the effective aperture region is obtained;

[0010] S3: The center line position of the pineapple leaf fiber is set as the axis, and the two-dimensional image plane is the corresponding cross section, a cylindrical coordinate system is established, and for each screened effective aperture region, the center point coordinates of each aperture inside are obtained by two-dimensional coordinate transformation; for each aperture center point coordinate, the radial distance from the fiber center line is calculated, the radial distance is added to the corresponding aperture radius, and compared with the equivalent radius of the single fiber branch region corresponding to the aperture region, if it is greater than the equivalent radius of the single fiber branch region, the aperture center point coordinate is excluded;

[0011] S4: Project the screened two-dimensional aperture center point coordinates into the cylindrical coordinate system to generate three-dimensional coordinates of the aperture center point, and based on the three-dimensional coordinates of the aperture center point, reconstruct the three-dimensional volume structure of the region occupied by each aperture by using the equivalent cylindrical model, and obtain the three-dimensional structure of the fiber internal aperture.

[0012] Further, the fiber center line, each fiber branch and the corresponding aperture region are identified, and the specific steps are as follows:

[0013] Collect the scanning electron microscope image of the cross section of the pineapple leaf fiber, set the background segmentation threshold, and assign the gray value of the fiber center line, each single fiber branch region and the corresponding aperture region above the background segmentation threshold to 255 and mark it as the foreground region, and assign the gray value of the pixel point region below the background segmentation threshold to 0 and mark it as the background region, finally generate a background segmentation image retaining the fiber center line, each single fiber and aperture region, wherein each fiber branch has one aperture region, and each aperture region contains multiple apertures;

[0014] Perform connected region analysis on the background segmentation image, identify all adjacent pixels marked as 255 as connected regions, count the number of pixels in each connected region, set a pixel number threshold, and mark the connected regions higher than the pixel number threshold as fiber branches and the connected regions lower than the pixel number threshold as aperture regions;

[0015] The connected regions marked as fiber branches are subjected to topological skeleton extraction by Zhang-Suen algorithm, and the edge pixels are gradually stripped off, and the single-pixel-width center line is retained as the fiber center line;

[0016] The specific process of the Zhang-Suen algorithm is as follows:

[0017] For each foreground pixel point, the eight adjacent pixels around it are taken as neighborhood pixels in a clockwise direction, and are numbered as above, right above, right, right below, below, left below, left, and left above in a clockwise direction. One round of iteration is performed, and the specific process is as follows:

[0018] The total number of white pixels in the eight adjacent pixels of each pixel point is counted, and is compared with the set number threshold. The pixel points not within the number threshold are filtered out. It is checked whether the number of black and white alternations of the neighborhood pixels is one. If not, the pixel points are deleted. A horizontal constraint is established. It is checked whether the left, lower, and right directions of the pixel points meeting the condition are not white at the same time. Then, it is checked whether the lower, right, and upper directions are not white at the same time. If the conditions are not met, the pixels on the boundary in the up-down direction are deleted, and an updated foreground image is generated.

[0019] Based on the updated foreground image, two rounds of iteration are performed, and the specific steps are as follows:

[0020] The neighborhood pixels of each foreground pixel point are re-labeled. The total number of white pixels in the eight adjacent pixels of each pixel point is counted, and is compared with the set number threshold. The pixel points not within the number threshold are filtered out. It is checked whether the number of black and white alternations of the neighborhood pixels is one. If not, the pixel points are deleted. It is checked whether the left, lower, and upper directions of the pixel points meeting the condition are not white at the same time. Then, it is checked whether the left, right, and upper directions are not white at the same time. If the conditions are not met, the pixels on the boundary in the left-right direction are deleted, and an image after two rounds of iteration is generated.

[0021] One round of iteration and two rounds of iteration are repeatedly performed until no pixel is deleted, a skeleton image of a fiber center line is formed, a background segmentation image of an original fiber branch is logically subtracted from the skeleton image, a single fiber branch region is generated, the image classified by the fiber center line, each single fiber, and the aperture region is re-labeled in terms of pixel points. The gray value of the center line pixel is marked as 255. The gray value of the pixel of the fiber branch region is marked as 128. The gray value of the aperture region is marked as 64. An updated second image is obtained.

[0022] Further, the equivalent circle method is used to calculate the equivalent diameter of each single fiber branch and the diameter of each single aperture, and the specific steps are as follows:

[0023] Based on the second image, the number of pixel points of each fiber branch region and each single aperture is counted. The equivalent circle method is used to calculate the equivalent diameter of each single fiber branch and the diameter of each single aperture.

[0024] The equivalent circle method involves multiplying the area of ​​a single pixel by the number of pixels in each fiber branch region and the number of pixels in a single aperture, then dividing the total area of ​​the obtained pixels by pi and taking the square root, and multiplying the result by 2 to obtain the equivalent diameter of each fiber branch region and the diameter of a single aperture.

[0025] Further, to obtain the effective aperture region, the specific steps are as follows:

[0026] Calculate the area of ​​each pore within its pore region, sum them up to obtain the area of ​​the corresponding pore region, convert the diameter of each pore obtained by the equivalent circle method into the corresponding circular area to obtain the total area, and then divide it by the total area of ​​its corresponding fiber branch region to calculate the area density of the pore region.

[0027] Calculate the expected value and standard deviation of the area density of the pore region corresponding to each fiber branch region in all the second images. Set the screening conditions. Only pore regions with area density between the expected value minus a fixed multiple of the standard deviation and the expected value plus that multiple of the standard deviation are considered valid. Valid pore regions are obtained through such screening.

[0028] Furthermore, the coordinates of the center point of each aperture within it are obtained using a two-dimensional coordinate transformation. The specific steps are as follows:

[0029] For each fiber branch region after screening, calculate the center point coordinates of each aperture. Specifically, take the average of the horizontal and vertical coordinates of all pixels in each aperture, and the resulting coordinates are the two-dimensional center point coordinates of each aperture.

[0030] Furthermore, the coordinates of the effective aperture center point are obtained, specifically:

[0031] For the center point coordinates of each aperture, the square root of the square of the horizontal and vertical coordinates is obtained by using the Pythagorean theorem to obtain the radial distance from the fiber centerline.

[0032] Add the radial distance to the corresponding aperture radius to obtain the range value of the aperture. Compare this range value with the equivalent radius of a single fiber branch region corresponding to the aperture region. If it is greater than the equivalent radius of a single fiber branch region, it means that the center point of the aperture is outside the effective range of the fiber branch region, and the coordinates of the center point of the aperture are excluded. Otherwise, the center point of the aperture is considered a valid point.

[0033] Furthermore, the two-dimensional coordinates of the aperture center point are projected onto a cylindrical coordinate system to generate the three-dimensional coordinates of the aperture center point, specifically as follows:

[0034] According to empirical formulas in materials science, the depth of an aperture is directly proportional to its diameter. Multiplying the diameter of the aperture by the depth-to-diameter ratio coefficient yields the depth at the center point of the aperture.

[0035] The two-dimensional coordinates of the center point of each aperture are projected onto the surface of a cylinder. The deviation angle of the center point of the aperture in the cylindrical coordinate system is obtained by calculating the arctangent value of the ordinate divided by the abscissa. Finally, the three-dimensional coordinates of each center point of the aperture are generated, including radial coordinates, angular coordinates and depth.

[0036] Furthermore, based on the three-dimensional coordinates of the aperture center point, an equivalent cylindrical model is used to reconstruct the three-dimensional volume structure of the region occupied by each aperture. The specific steps are as follows:

[0037] Based on the three-dimensional coordinates of the center point of the aperture, the diameter of each aperture is converted into the radius. Combined with the volume calculation formula of a cylinder, the depth of the aperture is used as the height of the cylinder, and the radius of the aperture is used as the base to calculate the volume of the cylinder and obtain the three-dimensional structure of the aperture inside the fiber.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] This invention proposes a secondary processing method based on scanning electron microscope images. Through background segmentation, region analysis, and 3D reconstruction, it can accurately extract the centerline, branch regions, and pore size features of pineapple leaf fibers. This method ensures high-precision analysis of fibers and pore sizes through a series of steps including pixel statistics, area normalization, and radial distance calculation. The introduction of a radial distance and equivalent radius comparison mechanism effectively eliminates edge artifacts and invalid pores, improving the purity of effective pore size data. Furthermore, by utilizing coordinate transformation in a cylindrical coordinate system, the data of the pore center point can be effectively projected into 3D space, thereby achieving 3D reconstruction of the pore size and accurately restoring the spatial distribution characteristics of pores within the fiber. This provides a reliable geometric basis for predicting material properties, enhances the depth and breadth of pineapple leaf fiber research, and promotes a comprehensive understanding of the microstructure and properties of natural materials. Attached Figure Description

[0040] Fig. 1 This is a schematic diagram of the overall method flow of the present invention;

[0041] Fig. 2 This is a schematic diagram showing the relationship between the depth of the center point of the aperture and the aperture volume. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0043] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0044] Example:

[0045] Please see Figs. 1-2 The present invention provides a technical solution:

[0046] The method for three-dimensional reconstruction of the microstructure of pineapple leaf fibers includes the following steps:

[0047] S1: Acquire scanning electron microscope images of pineapple leaf cross sections and label them as the first image. Perform background segmentation on the first image based on a preset segmentation threshold, identify the fiber centerline, each fiber branch region and the corresponding pore size region, and label the identified image as the second image.

[0048] The specific steps for identifying the fiber centerline, each fiber branch, and the corresponding aperture region are as follows:

[0049] Scanning electron microscope images of pineapple leaf fiber cross sections were acquired. After setting a background segmentation threshold, the gray values ​​of fiber center lines, individual fiber branch regions, and corresponding pore regions above the background segmentation threshold were assigned to 255 and marked as foreground regions. The gray values ​​of pixel regions below the background segmentation threshold were assigned to 0 and marked as background regions. Finally, a background segmentation image was generated that retained fiber center lines, individual fibers, and pore regions. Each fiber branch contained one pore region, and each pore region contained multiple pores.

[0050] Connectivity analysis is performed on the background segmentation image. All adjacent pixels marked as 255 are identified as connected regions. The number of pixels in each connected region is counted. A pixel count threshold is set. Connected regions with a pixel count higher than the threshold are marked as fiber branches, and connected regions with a pixel count lower than the threshold are marked as aperture regions.

[0051] The Zhang-Suen algorithm is used to extract the topological skeleton of the connected regions marked as fiber branches, gradually stripping away edge pixels and retaining the center line with a single pixel width as the fiber center line;

[0052] The specific process of the Zhang-Suen algorithm is as follows:

[0053] For each foreground pixel, its eight surrounding pixels are designated as neighboring pixels in a clockwise direction, and numbered clockwise as top, top right, right, bottom right, bottom, bottom left, left, top left, and so on. This process is repeated once. The specific steps are as follows:

[0054] Count the total number of white pixels among the eight neighboring pixels of each pixel and compare it with the set threshold. Filter out pixels that are not within the threshold. Check if the number of black and white alternations of neighboring pixels is 1. If not, delete them. Establish horizontal constraints and check if the left, bottom, and right directions of the pixels that meet the conditions are not white at the same time. Then check if the bottom, right, and top directions are not white at the same time. If the conditions are not met, delete the pixels at the top and bottom boundaries and generate the updated foreground image.

[0055] Based on the updated foreground image, a second round of iterations is performed, with the following specific steps:

[0056] Re-label the neighboring pixels of each foreground pixel, count the total number of white pixels in the 8 neighboring pixels of each pixel, and compare it with the set number threshold. Filter out pixels that are not within the number threshold. Check if the number of black and white alternations of the neighboring pixels is 1. If not, delete them. Check if the left, bottom, and top directions of the pixels that meet the conditions are not white at the same time. Then check if the left, right, and top directions are not white at the same time. If the conditions are not met, delete the pixels at the boundaries of the left and right directions. Generate the image after two rounds of iteration.

[0057] Repeat the first and second iterations until no pixels are deleted, forming a skeleton image of the fiber centerline. Perform logical subtraction between the original fiber branch background segmentation image and the skeleton image to generate each individual fiber branch region. Re-label the pixels of the image after classifying the fiber centerline, each individual fiber, and the aperture region. Label the gray values ​​of the centerline pixels as 255, the fiber branch region pixels as 128, and the aperture region pixels as 64 to obtain the updated second image.

[0058] In the above process, the background segmentation threshold is set by traversing all pixels of the image, counting the number of pixels at each gray level, and generating a gray-level histogram. Lower gray values ​​are represented by dense peaks on the left side of the histogram, and higher gray values ​​are represented by peaks on the right side of the histogram. The gray-level distribution of the background and the foreground forms two separate peaks with a clear trough in the middle. By directly observing the histogram, the lowest point between the two peaks is found, and the corresponding gray value is the background segmentation threshold.

[0059] By analyzing connected components, adjacent foreground pixels, i.e., regions with grayscale values ​​marked as 255, are divided into independent regions. Fiber branches and aperture regions are distinguished based on pixel count thresholds: a higher pixel count threshold filters out large continuous regions, i.e., fiber branches, avoiding misidentification of small apertures as fibers; a lower pixel count threshold captures small regions, i.e., individual apertures, ensuring the independence and statistical accuracy of apertures. Through size filtering, the structural features of fibers and apertures are effectively separated.

[0060] The method for setting the pixel count threshold here is to analyze all foreground regions after connecting regions, count the number of pixels in each region, generate a histogram of pixel count distribution, observe the distribution range of fiber branches and aperture, and set the pixel count threshold in the intermediate transition range by comparing the distribution ranges of the two.

[0061] The Zhang-Suen algorithm is used to extract the topological skeleton of the connected regions marked as fiber branches. In the first iteration, the upper and lower boundary pixels are deleted by horizontal constraints, and in the second iteration, the left and right boundary pixels are deleted by vertical constraints. The fiber skeleton is gradually refined. By checking the number of black and white alternations in the neighborhood, the continuity of the skeleton is ensured to avoid breakage or forking. After stripping the edge pixels, the single-pixel width of the fiber center line is retained to accurately describe the spatial direction of the fiber center line.

[0062] Logical subtraction is performed between the original fiber branch image and the skeleton image to separate the fiber peak region. Classification visualization is achieved through grayscale recalibration. The pixel of the fiber centerline is marked as 255 to highlight the geometric center of the fiber and provide a reference for subsequent 3D coordinate projection. The grayscale value of the pixel in the fiber branch region is marked as 128, and the grayscale value of the pore region is marked as 64. The structure type is intuitively distinguished by the grayscale difference, which facilitates the subsequent calculation of pore density. Logical subtraction can accurately preserve the true boundary of the fiber branch and avoid excessive cropping of the original region by the skeleton extraction.

[0063] The specific steps of logical subtraction are as follows: perform pixel-level set difference operation between the background segmentation image and the single-pixel skeleton image generated by the Zhang-Suen algorithm, and subtract the centerline pixels in the skeleton image that have been marked with a value of 255, thereby generating a background segmentation image that only retains the main area of ​​fiber branches, i.e., without the centerline. Specifically, if a pixel in the background segmentation image is marked with a value of 255 and a pixel in the skeleton image is marked with a value of 0, then the pixel is marked with a value of 255; if a pixel in the skeleton image is marked with a value of 255, then it is marked with a value of 0.

[0064] S2: Based on the second image, count the number of pixels in each fiber branch region and the number of individual pores within it. Calculate the equivalent diameter of each fiber branch and the diameter of each individual pore using the equivalent circle method. Based on the individual pore diameter, calculate the area density of the corresponding pore region using area normalization. Calculate the expected value and standard deviation of the pore region corresponding to each fiber branch region in all the second images. Set filtering conditions to filter the area density and obtain effective pore regions. The specific steps for calculating the equivalent diameter of each fiber branch and the diameter of each individual pore using the equivalent circle method are as follows:

[0065] The equivalent diameter of a single fiber branch and the diameter of a single aperture are calculated using the equivalent circle method. The specific steps are as follows:

[0066] Based on the second image, the number of pixels in each fiber branch region and a single aperture is counted, and the equivalent diameter of a single fiber branch and the diameter of a single aperture are calculated using the equivalent circle method.

[0067] The equivalent circle method involves multiplying the area of ​​a single pixel by the number of pixels in each fiber branch region and the number of pixels in a single aperture, then dividing the total area of ​​the obtained pixels by pi and taking the square root, and multiplying the result by 2 to obtain the equivalent diameter of each fiber branch region and the diameter of a single aperture.

[0068] In the above process, the actual shape of fiber branches and apertures is usually irregular, making it difficult and subjective to directly measure their geometric diameter. The equivalent circle method provides a unified and simplified characterization index by converting pixel regions of arbitrary shapes into equivalent circular diameters, eliminating the interference of morphological complexity on subsequent parameter extraction, and providing a unified diameter value that is independent of direction, which facilitates quantitative comparison across regions. The equivalent circle method is based on the overall area, is not sensitive to local edge noise, and is more robust.

[0069] The equivalent diameter of a single fiber branch and the diameter of a single aperture are calculated using the equivalent circle method. The specific formula is as follows:

[0070] Based on the second image, the number of pixels in each fiber branch region and for each individual aperture is counted. The equivalent diameter of a single fiber branch and the diameter of a single aperture are then calculated using the equivalent circle method.

[0071]

[0072] in, Indicates the first The number of pixels per aperture; Indicates the first One aperture diameter; This indicates the area of ​​a preset single pixel; Indicates the aperture index number;

[0073]

[0074] in, Indicates the first The equivalent diameter of a fiber branch; Indicates the first The number of pixels in each fiber branch; This indicates the fiber branch index number.

[0075] The specific steps for obtaining the effective aperture region are as follows:

[0076] Calculate the area of ​​each pore within its pore region, sum them up to obtain the area of ​​the corresponding pore region, convert the diameter of each pore obtained by the equivalent circle method into the corresponding circular area to obtain the total area, and then divide it by the total area of ​​its corresponding fiber branch region to calculate the area density of the pore region.

[0077] Calculate the expected value and standard deviation of the area density of the pore region corresponding to each fiber branch region in all the second images. Set the screening conditions. Only pore regions with area density between the expected value minus a fixed multiple of the standard deviation and the expected value plus that multiple of the standard deviation are considered valid. Valid pore regions are obtained through such screening.

[0078] In the above process, the density of effective pore size is concentrated around the mean, which is approximately a normal distribution. The screening condition is set to filter out pore size regions whose area density deviates significantly from the normal range, so as to avoid the interference of extreme values ​​on the overall statistical results. Only pore size regions consistent with the global density distribution are retained. The screening condition of expected value ± fixed multiple standard deviation aims to eliminate anomalies, ensure consistency and improve computational robustness.

[0079] The specific formula for obtaining the effective aperture region is as follows:

[0080] Based on a single aperture diameter, the area density of its corresponding aperture region is calculated using area normalization:

[0081]

[0082] in,

[0083] ,

[0084] Indicates the first Area density of each aperture region; Indicates the first Total number of apertures in each aperture region; Indicates the aperture region index number; Indicates belonging to the The set of aperture regions of the fiber branches;

[0085] Calculate the expected value and standard deviation of the aperture regions in all second images, set filtering criteria to filter by area density, and obtain the effective aperture regions:

[0086]

[0087] in, This represents the expected value of the aperture region for all second images; This represents the standard deviation of the aperture region for all second images; This represents the confidence coefficient.

[0088] S3: Set the centerline position of the pineapple leaf fiber to... The axis, the two-dimensional image plane is the corresponding For the cross-section, establish a cylindrical coordinate system. For each effective pore area after screening, use two-dimensional coordinate transformation to obtain the coordinates of the center point of each pore within it. For the center point coordinates of each pore, calculate its radial distance to the fiber centerline, add the radial distance to the corresponding pore radius, and compare it with the equivalent radius of the single fiber branch area corresponding to that pore area. If it is greater than the equivalent radius of the single fiber branch area, then exclude the center point coordinates of that pore.

[0089] The specific steps for obtaining the center point coordinates of each aperture within it using two-dimensional coordinate transformation are as follows:

[0090] For each fiber branch region after screening, calculate the center point coordinates of each aperture. Specifically, take the average of the horizontal and vertical coordinates of all pixels in each aperture, and the resulting coordinates are the two-dimensional center point coordinates of each aperture.

[0091] To obtain the coordinates of the center point of the effective aperture, specifically:

[0092] For the center point coordinates of each aperture, the square root of the square of the horizontal and vertical coordinates is obtained by using the Pythagorean theorem to obtain the radial distance from the fiber centerline.

[0093] Add the radial distance to the corresponding aperture radius to obtain the range value of the aperture. Compare this range value with the equivalent radius of a single fiber branch region corresponding to the aperture region. If it is greater than the equivalent radius of a single fiber branch region, it means that the center point of the aperture is outside the effective range of the fiber branch region, and the coordinates of the center point of the aperture are excluded. Otherwise, the center point of the aperture is considered a valid point.

[0094] The coordinates of the center point of the effective aperture are obtained using the following formula:

[0095] For the center point coordinates of each aperture, calculate its radial distance to the fiber centerline:

[0096]

[0097] in, Indicates the first The x-coordinate of the center point of each aperture; Indicates the first The ordinate of the center point of each aperture; Indicates the first The radial coordinates of the cylindrical coordinate system of the center point of each aperture;

[0098] Add the radial distance to the corresponding aperture radius and compare it with the equivalent radius of a single fiber branch region corresponding to that aperture region. If it is greater than the equivalent radius of a single fiber region, then exclude the center point coordinates of that aperture. The specific formula is as follows:

[0099]

[0100] in,

[0101]

[0102]

[0103] in, Indicates the first One aperture radius; Indicates the first The equivalent radius of the branched region of the fiber.

[0104] S4: Project the selected two-dimensional pore center point coordinates onto a cylindrical coordinate system to generate the three-dimensional coordinates of the pore center point. Based on the three-dimensional coordinates of the pore center point, use an equivalent cylindrical model to reconstruct the three-dimensional volume structure of the area occupied by each pore, and obtain the three-dimensional structure of the pores inside the fiber.

[0105] The method describes projecting the two-dimensional coordinates of the aperture center point onto a cylindrical coordinate system to generate the three-dimensional coordinates of the aperture center point, specifically as follows:

[0106] According to empirical formulas in materials science, the depth of an aperture is directly proportional to its diameter. Multiplying the diameter of the aperture by the depth-to-diameter ratio coefficient yields the depth at the center point of the aperture.

[0107] The two-dimensional coordinates of the center point of each aperture are projected onto the surface of a cylinder. The deviation angle of the center point of the aperture in the cylindrical coordinate system is obtained by calculating the arctangent value of the ordinate divided by the abscissa. Finally, the three-dimensional coordinates of each center point of the aperture are generated, including radial coordinates, angular coordinates and depth.

[0108] The three-dimensional coordinates of the aperture center point are generated by projecting the two-dimensional coordinates onto a cylindrical coordinate system. The specific formula is as follows:

[0109] According to empirical formulas in materials science, the depth of an aperture is directly proportional to its diameter.

[0110]

[0111] in, , represents the depth-to-diameter ratio coefficient; Indicates the first Depth of the center point of each aperture;

[0112] In the above formula, The physical meaning is the longitudinal expansion scale of the pores along the fiber centerline; measured by the depth-to-diameter ratio coefficient. The correlation between pore diameter and depth reflects the distribution characteristics of pores in three-dimensional space; It directly determines the longitudinal depth, reflecting the direct proportional relationship between aperture size and longitudinal extension; As an empirical coefficient, the ratio of pore depth to diameter is adjusted;

[0113] Center point of the two-dimensional aperture Projected onto the cylindrical surface:

[0114]

[0115] in, Indicates the first The angular coordinates of the center points of each aperture are used to generate the three-dimensional coordinates of each aperture center. .

[0116] In the above formula, Physically, it refers to the circumferential distribution of pores within the fiber cross-section. Mapping two-dimensional planar coordinates to three-dimensional cylindrical coordinates preserves the spatial orientation information of the pores, ensuring a true reconstruction of the pore topology. It reflects the direct influence of the original planar coordinates on the three-dimensional azimuth angle. Changes in the ratio of the two-dimensional coordinates directly lead to dynamic adjustment of the azimuth angle, preserving the continuity of the spatial distribution.

[0117] Based on the three-dimensional coordinates of the center point of the aperture, the diameter of each aperture is converted into the radius. Combined with the volume calculation formula of a cylinder, the depth of the aperture is used as the height of the cylinder, and the radius of the aperture is used as the base to calculate the volume of the cylinder and obtain the three-dimensional structure of the aperture inside the fiber.

[0118] The three-dimensional coordinates based on the aperture center point are used to reconstruct the three-dimensional volume structure of the region occupied by each aperture using an equivalent cylindrical model. The specific formula is as follows:

[0119] Based on the three-dimensional coordinates of the pore center point, the three-dimensional volume of each pore is constructed to obtain the three-dimensional structure of the pores inside the fiber:

[0120]

[0121] in, Indicates the first The volume at the center point of each aperture.

[0122] In the above formula, the dependent variable The physical meaning is that a single pore is located along the axial direction inside the fiber. The spatial occupancy value under the combined action of direction and radial direction The radial component that determines the volume. The pore diameter is the axial extension length along the fiber centerline, which characterizes the longitudinal expansion scale of the volume. The two are multiplied together to form the core parameter for volume calculation. The square of the pore diameter will significantly increase the volume, and the increase in axial extension length will directly lead to linear volume growth.

[0123] In this embodiment, 10 are selected. A diagram showing the relationship between the depth of the aperture center point and the aperture volume was constructed, and the experimental data are shown in Table 1:

[0124] Table 1: Schematic diagram of the relationship between the depth of the aperture center point and the aperture volume

[0125]

[0126] As can be seen from the table above, when As the aperture depth increases, the actual value of the aperture volume also increases. The actual value of the aperture volume fluctuates within a small range around the theoretical value of the aperture volume, showing an overall positive correlation, which reflects the direct proportional relationship between aperture volume and aperture depth.

[0127] Depend on Fig. 2It can be seen that the actual values ​​of pore volume and pore depth are generally directly proportional, and fluctuate within a small range above and below the fitted curve composed of the theoretical values ​​of pore volume, which is consistent with the characteristics of the theoretical calculation method of pore volume.

[0128] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0129] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0130] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0131] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for three-dimensional reconstruction of the microstructure of pineapple leaf fibers, characterized by the following steps: include: S1: Acquire scanning electron microscope images of pineapple leaf cross sections and label them as the first image. Perform background segmentation on the first image based on a preset segmentation threshold, identify the fiber centerline, each fiber branch region and the corresponding pore size region, and label the identified image as the second image. S2: Based on the second image, count the number of pixels in each fiber branch region and the number of individual apertures inside, and use the equivalent circle method to calculate the equivalent diameter of a single fiber branch and the diameter of a single aperture. Based on the diameter of a single aperture, use area normalization to calculate the area density of the corresponding aperture region, and calculate the expectation and standard deviation of the aperture region corresponding to each fiber branch region in all the second images. Set filtering conditions to filter the area density to obtain the effective aperture region. S3: Set the centerline position of the pineapple leaf fiber as the axis, and the two-dimensional image plane as the corresponding cross-section. Establish a cylindrical coordinate system. For each effective pore area after screening, use two-dimensional coordinate transformation to obtain the center point coordinates of each pore within it. For the center point coordinates of each pore, calculate its radial distance to the fiber centerline. Add the radial distance to the corresponding pore radius and compare it with the equivalent radius of the single fiber branch area corresponding to that pore area. If it is greater than the equivalent radius of the single fiber branch area, then exclude the center point coordinates of that pore. S4: Project the selected two-dimensional pore center point coordinates onto a cylindrical coordinate system to generate the three-dimensional coordinates of the pore center point. Based on the three-dimensional coordinates of the pore center point, use an equivalent columnar model to reconstruct the three-dimensional volume structure of the area occupied by each pore to obtain the three-dimensional structure of the pores inside the fiber. The specific steps for obtaining the center point coordinates of each aperture within the aperture using two-dimensional coordinate transformation are as follows: For each fiber branch region after screening, calculate the center point coordinates of each aperture. Specifically, take the average of the horizontal and vertical coordinates of all pixels in each aperture, and the resulting coordinates are the two-dimensional center point coordinates of each aperture. The three-dimensional volume structure of the region occupied by each aperture is reconstructed using an equivalent cylindrical model based on the three-dimensional coordinates of the aperture center point. The specific steps are as follows: Based on the three-dimensional coordinates of the center point of the aperture, the diameter of each aperture is converted into the radius. Combined with the volume calculation formula of a cylinder, the depth of the aperture is used as the height of the cylinder, and the radius of the aperture is used as the base to calculate the volume of the cylinder and obtain the three-dimensional structure of the aperture inside the fiber.

2. The method for three-dimensional reconstruction of the microstructure of pineapple leaf fibers according to claim 1, characterized in that, The specific steps for identifying the fiber centerline, each fiber branch, and the corresponding aperture region are as follows: Scanning electron microscope images of pineapple leaf fiber cross sections were acquired. After setting a background segmentation threshold, the gray values ​​of fiber center lines, individual fiber branch regions, and corresponding pore regions above the background segmentation threshold were assigned to 255 and marked as foreground regions. The gray values ​​of pixel regions below the background segmentation threshold were assigned to 0 and marked as background regions. Finally, a background segmentation image was generated that retained fiber center lines, individual fibers, and pore regions. Each fiber branch contained one pore region, and each pore region contained multiple pores. Connectivity analysis is performed on the background segmentation image. All adjacent pixels marked as 255 are identified as connected regions. The number of pixels in each connected region is counted. A pixel count threshold is set. Connected regions with a pixel count higher than the threshold are marked as fiber branches, and connected regions with a pixel count lower than the threshold are marked as aperture regions. The Zhang-Suen algorithm is used to extract the topological skeleton of the connected regions marked as fiber branches, gradually stripping away edge pixels and retaining the center line with a single pixel width as the fiber center line; The specific process of the Zhang-Suen algorithm is as follows: For each foreground pixel, its eight surrounding pixels are designated as neighboring pixels in a clockwise direction, and numbered clockwise as top, top right, right, bottom right, bottom, bottom left, left, top left, and so on. This process is repeated once. The specific steps are as follows: Count the total number of white pixels among the eight neighboring pixels of each pixel and compare it with the set threshold. Filter out pixels that are not within the threshold. Check if the number of black and white alternations of neighboring pixels is 1. If not, delete them. Establish horizontal constraints and check if the left, bottom, and right directions of the pixels that meet the conditions are not white at the same time. Then check if the bottom, right, and top directions are not white at the same time. If the conditions are not met, delete the pixels at the top and bottom boundaries and generate the updated foreground image. Based on the updated foreground image, a second round of iterations is performed, with the following specific steps: Re-label the neighboring pixels of each foreground pixel, count the total number of white pixels in the 8 neighboring pixels of each pixel, and compare it with the set number threshold. Filter out pixels that are not within the number threshold. Check if the number of black and white alternations of the neighboring pixels is 1. If not, delete them. Check if the left, bottom, and top directions of the pixels that meet the conditions are not white at the same time. Then check if the left, right, and top directions are not white at the same time. If the conditions are not met, delete the pixels at the boundaries of the left and right directions. Generate the image after two rounds of iteration. Repeat the first and second iterations until no pixels are deleted, forming a skeleton image of the fiber centerline. Perform logical subtraction between the original fiber branch background segmentation image and the skeleton image to generate each individual fiber branch region. Re-label the pixels of the image after classifying the fiber centerline, each individual fiber, and the aperture region. Label the gray values ​​of the centerline pixels as 255, the fiber branch region pixels as 128, and the aperture region pixels as 64 to obtain the updated second image.

3. The method for three-dimensional reconstruction of the microstructure of pineapple leaf fibers according to claim 1, characterized in that, The equivalent diameter of a single fiber branch and the diameter of a single aperture are calculated using the equivalent circle method. The specific steps are as follows: Based on the second image, the number of pixels in each fiber branch region and a single aperture is counted, and the equivalent diameter of a single fiber branch and the diameter of a single aperture are calculated using the equivalent circle method. The equivalent circle method involves multiplying the area of ​​a single pixel by the number of pixels in each fiber branch region and the number of pixels in a single aperture, then dividing the total area of ​​the obtained pixels by pi and taking the square root, and multiplying the result by 2 to obtain the equivalent diameter of each fiber branch region and the diameter of a single aperture.

4. The method for three-dimensional reconstruction of the microstructure of pineapple leaf fibers according to claim 3, characterized in that, The specific steps for obtaining an effective aperture region are as follows: Calculate the area of ​​each pore within its pore region, sum them up to obtain the area of ​​the corresponding pore region, convert the diameter of each pore obtained by the equivalent circle method into the corresponding circular area to obtain the total area, and then divide it by the total area of ​​its corresponding fiber branch region to calculate the area density of the pore region. Calculate the expected value and standard deviation of the area density of the pore region corresponding to each fiber branch region in all the second images. Set the screening conditions. Only pore regions with area density between the expected value minus a fixed multiple of the standard deviation and the expected value plus that multiple of the standard deviation are considered valid. Valid pore regions are obtained through such screening.

5. The method for three-dimensional reconstruction of the microstructure of pineapple leaf fibers according to claim 1, characterized in that, To obtain the coordinates of the center point of the effective aperture, specifically: For the center point coordinates of each aperture, the square root of the square of the horizontal and vertical coordinates is obtained by using the Pythagorean theorem to obtain the radial distance from the fiber centerline. Add the radial distance to the corresponding aperture radius to obtain the range value of the aperture. Compare this range value with the equivalent radius of a single fiber branch region corresponding to the aperture region. If it is greater than the equivalent radius of a single fiber branch region, it means that the center point of the aperture is outside the effective range of the fiber branch region, and the coordinates of the center point of the aperture are excluded. Otherwise, the center point of the aperture is considered a valid point.

6. The method for three-dimensional reconstruction of the microstructure of pineapple leaf fibers according to claim 5, characterized in that, The filtered two-dimensional coordinates of the aperture center point are projected onto a cylindrical coordinate system to generate the three-dimensional coordinates of the aperture center point, specifically: According to empirical formulas in materials science, the depth of an aperture is directly proportional to its diameter. Multiplying the diameter of the aperture by the depth-to-diameter ratio coefficient yields the depth at the center point of the aperture. The two-dimensional coordinates of the center point of each aperture are projected onto the surface of a cylinder. The deviation angle of the center point of the aperture in the cylindrical coordinate system is obtained by calculating the arctangent value of the ordinate divided by the abscissa. Finally, the three-dimensional coordinates of each center point of the aperture are generated, including radial coordinates, angular coordinates and depth.

Images