A method for detecting the wear resistance of skateboard wheels

Abstract

The present invention discloses a method for detecting the wear resistance of a skateboard wheel, belonging to the field of image recognition technology. The method comprises the following steps: obtaining a grayscale image of the pulley surface; obtaining a grayscale histogram of the grayscale image; obtaining an enhanced grayscale image corresponding to each first segmentation point; obtaining the clarity of the closed edge line in the enhanced grayscale image; using the enhanced grayscale image corresponding to the maximum quality evaluation value as the optimal grayscale image; and obtaining the wear resistance of the pulley surface based on the density of the closed areas in the optimal grayscale image and the uniformity of the areas of all closed areas in the optimal grayscale image. The present invention determines whether there is a large defect location on the skateboard wheel surface by analyzing the relationship between the edge clarity of the pits in the enhanced image and the grayscale uniformity of the pixels within the closed edges, thereby analyzing the uniformity of the wear degree of the pulley, thereby evaluating the wear resistance of the pulley.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a method for detecting the wear resistance of a skateboard wheel. Background Art

[0002] After pulley production, new pulleys need to be sampled and tested for wear resistance. The wear resistance of the pulley surface is primarily measured using a wear tester, where the wear resistance is measured by measuring the roughness of the pulley surface at different times. Prior art methods for detecting surface wear typically analyze the wear degree of the pulley by simply preprocessing the image and then analyzing the number and size of pits on the pulley surface. However, because the pits and the pulley surface have similar colors, the number and area of ​​the pits obtained during image segmentation are inaccurate, resulting in inaccurate results for pulley wear resistance testing. To clearly segment pit defects, those skilled in the art typically enhance the image. Histogram equalization is typically used to adjust the grayscale values ​​using a cumulative function to enhance contrast. However, this enhancement process reduces the grayscale level of the transformed image, causing weak edge details in the image to disappear, making it difficult to accurately detect pit edges in the pulley surface image. Summary of the Invention

[0003] In order to address the deficiencies in the prior art, the present invention provides a method for detecting the wear resistance of a skateboard wheel. The method determines whether there is a large defect location on the surface of the skateboard wheel by enhancing the relationship between the edge clarity of the pits in the image and the grayscale uniformity of the pixels inside the closed edge, analyzes the uniformity of the wear degree of the pulley, and thus evaluates the wear resistance of the pulley.

[0004] The object of the present invention is to provide a method for detecting the wear resistance of a skateboard wheel, comprising the following steps:

[0005] Acquire a grayscale image of the pulley surface; and acquire a grayscale histogram of the grayscale image;

[0006] Each grayscale value in the grayscale image is used as a first segmentation point; the grayscale value with the highest frequency in the grayscale histogram is used as a second segmentation point; a piecewise linear transformation is performed on the grayscale value in the grayscale image according to each first segmentation point and the second segmentation point to obtain an enhanced grayscale image corresponding to each first segmentation point;

[0007] Perform edge detection on the enhanced grayscale image to obtain closed edge lines and closed areas enclosed by closed edge lines, as well as endpoints of non-closed edge lines in the enhanced grayscale image; obtain edge continuity in the enhanced grayscale image based on the average of distances between all adjacent endpoints and the number of endpoints; obtain clarity of closed edge lines in the enhanced grayscale image based on the edge continuity in the enhanced grayscale image and the contrast of closed edge lines in the enhanced grayscale image;

[0008] Obtain a quality evaluation value of the enhanced grayscale image corresponding to each first segment point based on the grayscale uniformity of pixels within all closed areas of the enhanced grayscale image and the clarity of closed edge lines in the enhanced grayscale image; and take the enhanced grayscale image corresponding to the maximum quality evaluation value as the optimal grayscale image;

[0009] The wear resistance of the pulley surface is obtained according to the density of the closed regions in the optimal grayscale image and the uniformity of the areas of all the closed regions in the optimal grayscale image.

[0010] In one embodiment, the contrast of the closed edge lines in the enhanced grayscale image is obtained by following the steps below:

[0011] Obtaining the contrast of each closed edge line in the enhanced grayscale image according to the grayscale difference between the edge pixel point and its neighboring pixel points on each closed edge line in the enhanced grayscale image;

[0012] The mean of the contrasts of all closed edge lines is used as the contrast of the closed edge lines in the enhanced grayscale image.

[0013] In one embodiment, the density of closed areas in the optimal grayscale image is obtained according to the following steps:

[0014] The entropy value of the distance between the center of each closed region in the optimal grayscale image and the center of its adjacent closed region is used as the density of the closed region in the optimal grayscale image.

[0015] In one embodiment, the uniformity of the areas of all closed regions in the optimal grayscale image is obtained according to the following steps:

[0016] The variance of the areas of all closed regions is taken as the uniformity of the areas of all closed regions in the optimal grayscale image.

[0017] In one embodiment, the method for obtaining the grayscale uniformity of pixels in all closed areas in the enhanced grayscale image includes:

[0018] The entropy value of the grayscale value of the pixel points in all closed areas in the enhanced grayscale image is used as the grayscale uniformity of the pixel points in all closed areas in the enhanced grayscale image.

[0019] In one embodiment, the enhanced grayscale image corresponding to each first segmentation point is obtained according to the following steps:

[0020] Performing a piecewise linear transformation on the grayscale value in the grayscale image according to each first segmentation point and the second segmentation point, obtaining the transformed grayscale value, and obtaining a grayscale image corresponding to the transformed grayscale value;

[0021] The grayscale image corresponding to the transformed grayscale value is used as the enhanced grayscale image corresponding to each first segmentation point.

[0022] The beneficial effects of the present invention are as follows: a method for detecting the wear resistance of a skateboard wheel provided by the present invention is provided, which enhances the grayscale image of the pulley surface based on a piecewise linear change method by selecting different segmentation points, thereby improving the contrast between the edge information and the background in the grayscale image, and then based on the contrast between the edge information and the background in the enhanced grayscale image, and the grayscale uniformity of the pixel points in the closed area, obtains the quality evaluation value of the enhanced grayscale image corresponding to each first segmentation point and the second segmentation point, and constructs an evaluation of the quality of the grayscale image through the features of the edges of the pits in the enhanced image, thereby obtaining the first segmentation point corresponding to the highest quality evaluation value and the corresponding enhanced grayscale image, so that the feature distribution in the obtained enhanced image is more obvious, and in the subsequent wear performance detection process of the pulley surface, the accuracy of detecting the wear degree of the pulley is higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0024] Figure 1 The present invention is a flowchart of the overall steps of a method for detecting the wear resistance of a skateboard wheel according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0026] The main purpose of the present invention is to use image recognition technology to obtain the grayscale values ​​in the grayscale image through piecewise linear transformation of the image obtained by different segmentation points, thereby improving the contrast between edge information and background. By improving the edge contrast, the edge information in the image is analyzed and judged to obtain the degree of wear on the pulley surface, thereby improving the detection accuracy of the pulley wear resistance.

[0027] The scenario targeted by this invention is that after the pulley is produced industrially, the wear resistance of the pulley needs to be tested. The wear resistance of the pulley affects the service life of the pulley. The pulley is tested on a scooter wear tester, and the wear degree of the pulley is judged by machine vision after the test.

[0028] The present invention provides a method for detecting the wear resistance of a skateboard wheel, see Figure 1 As shown, the following steps are included:

[0029] S1. Obtain a grayscale image of the pulley surface; and obtain a grayscale histogram of the grayscale image;

[0030] In this embodiment, a wear test is performed by mounting the pulley to be tested on a wear testing machine, and a camera is installed on the testing machine. The camera captures the surface image of the pulley after the test, and the roughness of the pulley surface is evaluated based on the number and distribution of pits in the surface image after the test.

[0031] A high-resolution camera is mounted in front of the pulley on the testing machine, and light sources are installed behind and above the pulley workpiece to capture the surface image of the pulley workpiece. This surface image is then grayscale processed to obtain a grayscale image of the pulley surface, and a grayscale histogram of the grayscale image is obtained.

[0032] S2. Taking each grayscale value in the grayscale image as a first segmentation point; taking the grayscale value with the highest frequency in the grayscale histogram as a second segmentation point; performing a piecewise linear transformation on the grayscale values ​​in the grayscale image according to each first segmentation point and the second segmentation point, and obtaining an enhanced grayscale image corresponding to each first segmentation point;

[0033] It should be noted that when evaluating the pulley surface roughness based on the number and distribution of pits in the pulley surface grayscale image, a clear grayscale image is required to accurately evaluate the pulley surface roughness. Therefore, in this embodiment, the purpose of enhancing the pulley surface is mainly to make the edges of the pits in the obtained pulley surface image clearer. Because the Canny algorithm is difficult to detect weak edges in the image during edge detection, and the edges in the resulting image are not clear, image enhancement is required to improve edge contrast.

[0034] In this embodiment, piecewise linear enhancement is primarily used to segment the grayscale range in an image, thereby improving the contrast between different image regions, namely the background and target regions, and enhancing edge clarity, thereby improving the accuracy of pothole detection. During the enhancement process, the first and second segmentation points for initial enhancement are first selected. The second segmentation point during enhancement uses the grayscale value with the highest frequency in the grayscale histogram as the value of the second segmentation point. The first segmentation point is filtered based on the impact of different grayscale values ​​on the image. The minimum grayscale value in the grayscale image can be used as the value of the initial first segmentation point, and the initial first segmentation points are traversed with a step size of 1 until the maximum grayscale value in the grayscale image is reached. The grayscale image is enhanced by using different first and second segmentation points, and the optimal first segmentation point is selected based on the characteristic changes of the edges in the enhanced grayscale image.

[0035] Specifically, the enhanced grayscale image corresponding to each first segmentation point is obtained according to the following steps:

[0036] Performing a piecewise linear transformation on the grayscale value in the grayscale image according to each first segmentation point and the second segmentation point, obtaining the transformed grayscale value, and obtaining a grayscale image corresponding to the transformed grayscale value;

[0037] The grayscale image corresponding to the transformed grayscale value is used as the enhanced grayscale image corresponding to each first segmentation point.

[0038] It should be noted that image enhancement based on piecewise linear changes is an existing technology and will not be described in detail here.

[0039] S3. Perform edge detection on the enhanced grayscale image to obtain closed edge lines and closed areas enclosed by closed edge lines, as well as endpoints of non-closed edge lines in the enhanced grayscale image; obtain edge continuity in the enhanced grayscale image based on the average of distances between all adjacent endpoints and the number of endpoints in the enhanced grayscale image; obtain clarity of closed edge lines in the enhanced grayscale image based on the edge continuity in the enhanced grayscale image and the contrast of closed edge lines in the enhanced grayscale image;

[0040] It should be noted that when detecting pits in the resulting grayscale image, the edge detection of the pits is unclear, and multiple pits may be detected as one. For a good pulley surface, the edge gradient in the grayscale image is large, the edge continuity is good, and the pixel points in the closed area are evenly distributed. After enhancing the image by selecting appropriate segmentation points, if the enhanced segmentation points are not selected well, the edges of the pitted portion of the pulley surface are unclear. When segmenting the image, the resulting segmentation blocks may be composed of multiple pits, and the pixel points within the segmentation blocks are unevenly distributed. The clarity of the pit edge in the image and the uniformity of the pixels within the segmentation blocks represent the quality of the currently selected threshold. The higher the clarity, the better the segmentation point selection, and the smaller the clarity, the worse the segmentation point selection effect. When edge detection is performed on the resulting image using the Canny edge detection operator, the resulting edge curve is unclear, and weaker edges in the image are difficult to detect. This leads to inaccurate results when analyzing wear resistance based on the distribution and size of the pits. Therefore, the image quality is analyzed by the changes in the edges in the enhanced image corresponding to different first segmentation points, as follows:

[0041] First, edge detection is performed on the enhanced grayscale image based on the canny edge detection operator to obtain the closed edge lines and the closed areas enclosed by the closed edge lines in the enhanced grayscale image, as well as the endpoints of the non-closed edge lines; the edge continuity is judged based on the distribution of the distances between the endpoints. If the number of endpoints in the obtained image is The smaller the minimum distance L between the endpoints, the better the edge coherence in the corresponding image. The calculation formula for enhancing the edge coherence in the grayscale image is as follows:

[0042]

[0043] Where, Indicates enhancing the edge coherence in grayscale images; Represents the mean of the distances between all adjacent endpoints in the enhanced grayscale image; The first segment point is The number of endpoints in the enhanced grayscale image corresponding to the time; wherein, the mean value of the distance between all adjacent endpoints in the enhanced grayscale image is calculated as follows: , where Indicates the first The coordinates of the edge endpoints; Indicates the first The coordinates of the edge endpoints; Expressed as the total number of all endpoints in the enhanced grayscale image, Represents the mean of the distances between all adjacent endpoints in the enhanced grayscale image.

[0044] It should be noted that the fewer endpoints in the enhanced grayscale image and the smaller the distance between them, the better the edge continuity in the resulting enhanced grayscale image, and the better the detection of pit edges in the enhanced grayscale image. This corresponds to the fact that the edges in the enhanced grayscale image after Canny edge detection are closed regions, and no endpoints exist in closed regions. Furthermore, the fewer endpoints in the resulting enhanced grayscale image, the better the pit detection effect. Because Canny is less effective at detecting weak edges, the edges in the resulting enhanced grayscale image become disconnected, resulting in endpoints. The smaller the distance between the two separated endpoints, the better the pit edge detection effect.

[0045] Secondly, the contrast of closed edge lines in the enhanced grayscale image is obtained according to the following steps: the contrast of each closed edge line in the enhanced grayscale image is obtained based on the grayscale difference between the edge pixel and its neighboring pixels on each closed edge line in the enhanced grayscale image; and the average of the contrasts of all closed edge lines is used as the contrast of the closed edge lines in the enhanced grayscale image. The contrast of each closed edge line in the enhanced grayscale image is calculated as follows:

[0046]

[0047] Where, It represents the first closed edge in the enhanced grayscale image. The gray value of the edge point, Represented as the first The first edge point in the 8-neighborhood The gray value of a pixel, Represents the total number of pixels on the closed edge in the enhanced grayscale image, It is expressed as the mean of the difference between all edge pixels and neighboring pixels in the tth closed edge region in the enhanced grayscale image. The larger the value, the clearer the corresponding closed edge pixel point; according to the first segmentation point obtained The mean value of the edge contrast of all closed edges in the corresponding enhanced grayscale image is , and the mean value of the contrast of all closed edge lines is calculated as follows: ,in, The first segment point is The mean contrast of all closed edge lines in the corresponding enhanced grayscale image is taken as the first segmentation point. The contrast of closed edge lines in the grayscale image is enhanced correspondingly when ; T represents the number of closed edges.

[0048] It should be noted that the greater the mean value of the difference between all edge pixels and neighboring pixels in the tth closed edge region in the calculated enhanced grayscale image, the greater the contrast of the corresponding closed edge, and thus the clearer the corresponding closed edge.

[0049] Finally, according to the edge continuity in the enhanced grayscale image and the contrast of the closed edge line in the enhanced grayscale image, the clarity of the closed edge line in the enhanced grayscale image is obtained. The clarity calculation formula of the closed edge line in the enhanced grayscale image is as follows:

[0050]

[0051] Where, Indicates enhancing the edge coherence in grayscale images; The first segment point is The contrast of closed edge lines in the corresponding enhanced grayscale image; The first segment point is The clarity of closed edge lines in the corresponding enhanced grayscale image.

[0052] It should be noted that the smaller the distance between the edge pixels in the enhanced grayscale image, the greater the difference between the edge pixels and the neighboring pixels, and the clearer the edge. The corresponding enhanced edge coherence in the grayscale image The larger the value of The larger the value of , the better the segmentation edges in the enhanced grayscale image.

[0053] S4. Obtaining a quality evaluation value of the enhanced grayscale image corresponding to each first segment point based on the grayscale uniformity of pixels within all closed areas of the enhanced grayscale image and the clarity of closed edge lines in the enhanced grayscale image; and taking the enhanced grayscale image corresponding to the maximum quality evaluation value as the optimal grayscale image;

[0054] The method for obtaining the grayscale uniformity of pixels in all closed areas of the enhanced grayscale image includes:

[0055] The entropy of the grayscale values ​​of all pixels in the closed area of ​​the enhanced grayscale image is used as the grayscale uniformity of all pixels in the closed area of ​​the enhanced grayscale image. The grayscale uniformity calculation formula of all pixels in the closed area of ​​the enhanced grayscale image is as follows:

[0056]

[0057] Where, Represents the frequency of occurrence of the grayscale value of the i-th pixel in all closed areas of the enhanced grayscale image; represents a logarithmic function with base 2; The first segment point is The entropy value of the grayscale value of all pixels in the closed area of ​​the corresponding enhanced grayscale image is used as the grayscale uniformity of all pixels in the closed area of ​​the enhanced grayscale image. It should be noted that the more uniform the distribution of the grayscale values ​​of the pixels in the closed area is, the greater the probability that the closed area is a pit; if the grayscale value distribution in the closed area is more uneven, the entropy value of the calculated grayscale value will be higher. The larger the value, the more holes there are inside the closed area. The reason for this is that the edge gradient of the holes is small. During the canny edge detection process, the weak edge detection effect in the image is poor. By selecting different first segmentation points for judgment and analysis, the contrast between the edges and other areas in the grayscale image can be effectively enhanced. Therefore, the entropy value of the grayscale value of all pixels in the closed area of ​​the grayscale image is enhanced. The smaller it is, the better the effect of image edge detection will be.

[0058] It should be noted that the more uniform the pixel value distribution of all the pixels in the closed area is, the better the segmentation effect of the closed area in the corresponding enhanced grayscale image is, which means that the contrast between the segments in the enhanced grayscale image is better. Therefore, by The image is evaluated at the same time, and the higher the evaluation value of the image quality is, the better the effect of the corresponding selected segmentation points is.

[0059] There are still weak edges in the enhanced grayscale image that have not been detected. Therefore, the distribution of pixels in the closed area of ​​the enhanced grayscale image is uneven, which affects the judgment of the area of ​​the pits in the enhanced grayscale image. The quality of the image needs to be evaluated based on the clarity of the edges in the enhanced grayscale image and whether the distribution of pixels inside the closed area is uniform. The main factor affecting the image quality is the clarity of the edges in the image, so the clarity of the edges occupies a larger proportion.

[0060] To this end, in this embodiment, the quality evaluation value of the enhanced grayscale image corresponding to each first segment point is obtained based on the grayscale uniformity of the pixels in all closed areas in the enhanced grayscale image and the clarity of the closed edge lines in the enhanced grayscale image. The quality evaluation value of the enhanced grayscale image corresponding to each first segment point is calculated as follows:

[0061]

[0062] Where, The first segment point is The quality evaluation value of the corresponding enhanced grayscale image; The first segment point is The entropy value of the grayscale values ​​of all pixels in the closed area of ​​the corresponding enhanced grayscale image is the grayscale uniformity of all pixels in the closed area of ​​the enhanced grayscale image; The first segment point is The clarity of the closed edge line in the corresponding enhanced grayscale image; 、 The first segment point is The weight ratio corresponding to the grayscale uniformity and clarity in the corresponding enhanced grayscale image; in this embodiment, , The proportion of each part can be adjusted according to the needs; it should be noted that the advantage of evaluating the image based on the clarity of the closed edge line in the enhanced grayscale image and the grayscale uniformity of the pixels in all closed areas in the enhanced grayscale image is that the more uniform the grayscale value distribution of the pixels in all closed areas is, the better the segmentation effect of the closed area in the corresponding enhanced grayscale image is, which means that the contrast between the segmented blocks in the enhanced grayscale image is better; since the entropy value of the grayscale value of the pixels in all closed areas in the enhanced grayscale image is The smaller it is, the better the effect of image edge detection will be; the first segmentation point is The larger the clarity value of the closed edge line in the enhanced grayscale image, the better the segmented edge in the enhanced grayscale image. The image quality can be better evaluated by evaluating the grayscale uniformity and the clarity of the closed edge line. The higher the evaluation value, the clearer the edge information of the image. In addition, the main factor affecting the image quality is the clarity of the edge in the image, so the edge clarity setting The entropy value of the gray value is set Therefore, the quality evaluation of the enhanced grayscale image is more accurate according to the clarity of the closed edge line in the enhanced grayscale image and the grayscale uniformity of the pixels in all closed areas in the enhanced grayscale image at different proportions.

[0063] The quality evaluation values ​​of the enhanced grayscale images corresponding to different first segmentation points are calculated in sequence, and the enhanced grayscale image corresponding to the maximum quality evaluation value is taken as the optimal grayscale image; and the first segmentation point corresponding to the maximum quality evaluation value is taken as the optimal first segmentation point.

[0064] At this point, the grayscale values ​​in the grayscale image are subjected to a piecewise linear transformation using the calculated optimal first and second segmentation points. The resulting enhanced grayscale image serves as the optimal grayscale image. This optimal grayscale image is then used to detect wear resistance, improving the accuracy of subsequent analysis of the pulley's surface wear.

[0065] S5. Obtain the wear resistance of the pulley surface according to the density of the closed regions in the optimal grayscale image and the uniformity of the areas of all closed regions in the optimal grayscale image.

[0066] It should be noted that the wear resistance of the pulley surface is evaluated based on the area and density of the closed area in the optimal grayscale image. When the wear resistance of the pulley is good, after the surface of the pulley is tested by the wear tester, the better the wear resistance of the pulley, the more uniform the area of ​​the pits in the optimal grayscale image obtained and the more dense the distribution of the pits, indicating that the wear resistance of the pulley is good; the individual area of ​​the pits in the optimal grayscale image obtained is large, and the distribution between the divided blocks is uneven, indicating that the wear resistance of the pulley surface is poor, resulting in uneven size and uneven distribution of the wear pits on the pulley surface during the test.

[0067] Specifically, the density of the closed area in the optimal grayscale image is obtained according to the following steps:

[0068] The entropy value of the distance between the center of each closed region in the optimal grayscale image and the center of its adjacent closed region is used as the density of the closed region in the optimal grayscale image.

[0069] In this embodiment, the distance between the center points of the closed edge areas in the optimal grayscale image after performing the Canny algorithm edge detection on the optimal grayscale image is obtained to obtain the distance between the center points of the adjacent closed areas. , get the minimum value of the center distance of all adjacent closed areas and maximum value , and count the frequencies of occurrence at different distances The density of the closed area in the optimal grayscale image is characterized by the entropy of the coordinates obtained according to the frequency of the center point coordinate position of the closed area in the optimal grayscale image: The density calculation formula of the closed area in the optimal grayscale image is as follows:

[0070]

[0071] Where, The distance between the center of each closed region and the center of its adjacent closed region is Frequency of occurrence; 、 Respectively represent the minimum and maximum values ​​of the distances between the centers of all adjacent closed areas; the frequency of occurrence is obtained by counting the distances between the center points of adjacent segmented blocks in the optimal grayscale image; The entropy value representing the distance between the center of each closed area in the optimal grayscale image and the center of its adjacent closed area is the density of the closed area in the optimal grayscale image. The larger the value of , the more scattered the distribution of the holes in the image. The smaller the value, the denser the distribution of the pits.

[0072] Furthermore, the uniformity of the areas of all closed regions in the optimal grayscale image is obtained according to the following steps:

[0073] The variance of the area of ​​all closed regions is used as the uniformity of the area of ​​all closed regions in the optimal grayscale image. The uniformity of the area of ​​all closed regions in the optimal grayscale image is calculated as follows:

[0074]

[0075] Where, Represented as the optimal grayscale image The area of ​​a closed region, Expressed as the mean of the closed area in the optimal grayscale image, Expressed as the total number of closed regions in the optimal grayscale image, Represents the uniformity of the area of ​​all closed regions in the optimal grayscale image; the calculation of the uniformity of the area of ​​all closed regions in the optimal grayscale image is mainly based on the variance to find the uniformity of the distribution of closed regions, where The smaller the value, the more uniform.

[0076] Furthermore, the wear resistance of the pulley surface is obtained based on the density of the closed areas in the optimal grayscale image and the uniformity of the areas of all closed areas in the optimal grayscale image. The wear resistance of the pulley surface is calculated as follows:

[0077]

[0078] Where, The entropy value representing the distance between the center of each closed area in the optimal grayscale image and the center of its adjacent closed area is the density of the closed areas in the optimal grayscale image. Indicates the uniformity of the area of ​​all closed regions in the optimal grayscale image; It is expressed as the wear resistance of the pulley surface; the worse the wear resistance, the larger the pits produced on the pulley surface after the test; the denser the distribution of the pits, the more uniform the wear resistance of the pulley surface; the larger the area of ​​the enhanced segmented blocks and the more discrete the distribution, the worse the wear resistance of the corresponding pulley surface; the smaller the area of ​​the segmented blocks and the more uniform the distribution, the better the wear resistance of the pulley.

[0079] This embodiment can evaluate the wear resistance of the pulley surface based on the area and density of the closed areas in the optimal grayscale image. When the pulley has good wear resistance, after testing the pulley surface with a wear tester, the better the wear resistance of the pulley, the more uniform the area of ​​the pits in the optimal grayscale image and the denser the distribution of the pits, indicating that the pulley has good wear resistance. If the individual pits in the optimal grayscale image are large in area and unevenly distributed between the segments, it indicates that the pulley surface has poor wear resistance, resulting in uneven sizes and uneven distribution of wear pits on the pulley surface during the test. Therefore, the operator can judge the wear resistance of the pulley based on the wear resistance of the pulley surface.

[0080] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for detecting the wear resistance of a skateboard wheel, characterized in that: The following steps are involved: Acquire a grayscale image of the pulley surface; and acquire a grayscale histogram of the grayscale image; Each grayscale value in the grayscale image is used as the first segmentation point; the grayscale value with the highest frequency in the grayscale histogram is used as the second segmentation point; Performing a piecewise linear transformation on the grayscale value in the grayscale image according to each first segmentation point and the second segmentation point to obtain an enhanced grayscale image corresponding to each first segmentation point; Performing edge detection on the enhanced grayscale image to obtain closed edge lines and closed areas enclosed by the closed edge lines, as well as endpoints of non-closed edge lines in the enhanced grayscale image; Obtaining edge coherence in the enhanced grayscale image according to the mean of the distances between all adjacent endpoints and the number of endpoints in the enhanced grayscale image; Obtaining the clarity of the closed edge line in the enhanced grayscale image according to the edge continuity in the enhanced grayscale image and the contrast of the closed edge line in the enhanced grayscale image; Obtaining a quality evaluation value of the enhanced grayscale image corresponding to each first segment point based on the grayscale uniformity of pixels in all closed areas of the enhanced grayscale image and the clarity of closed edge lines in the enhanced grayscale image; The enhanced grayscale image corresponding to the maximum quality evaluation value is taken as the optimal grayscale image; The wear resistance of the pulley surface is obtained according to the density of the closed regions in the optimal grayscale image and the uniformity of the areas of all the closed regions in the optimal grayscale image.

2. The method for detecting the wear resistance of a skateboard wheel according to claim 1, wherein: The contrast of the closed edge line in the enhanced grayscale image is obtained by following the steps below: Obtaining the contrast of each closed edge line in the enhanced grayscale image according to the grayscale difference between the edge pixel point and its neighboring pixel points on each closed edge line in the enhanced grayscale image; The mean of the contrasts of all closed edge lines is used as the contrast of the closed edge lines in the enhanced grayscale image.

3. The method for detecting the wear resistance of a skateboard wheel according to claim 1, wherein: The density of the closed area in the optimal grayscale image is obtained according to the following steps: The entropy value of the distance between the center of each closed region in the optimal grayscale image and the center of its adjacent closed region is used as the density of the closed region in the optimal grayscale image.

4. The method for detecting the wear resistance of a skateboard wheel according to claim 1, wherein: The uniformity of the areas of all closed regions in the optimal grayscale image is obtained by following the steps below: The variance of the areas of all closed regions is taken as the uniformity of the areas of all closed regions in the optimal grayscale image.

5. The method for detecting the wear resistance of a skateboard wheel according to claim 1, wherein: The method for obtaining the grayscale uniformity of pixels in all closed areas of the enhanced grayscale image includes: The entropy value of the grayscale value of the pixel points in all closed areas in the enhanced grayscale image is used as the grayscale uniformity of the pixel points in all closed areas in the enhanced grayscale image.

6. The method for detecting the wear resistance of a skateboard wheel according to claim 1, wherein: The enhanced grayscale image corresponding to each first segmentation point is obtained according to the following steps: Performing a piecewise linear transformation on the grayscale value in the grayscale image according to each first segmentation point and the second segmentation point, obtaining the transformed grayscale value, and obtaining a grayscale image corresponding to the transformed grayscale value; The grayscale image corresponding to the transformed grayscale value is used as the enhanced grayscale image corresponding to each first segmentation point.

Images