An intelligent measurement method and system for wheel rim dimensions of supporting wheels

Abstract

The present application relates to the field of rim thickness measurement technology, and specifically to an intelligent rim size measurement method and system for a track roller, the method comprising: collecting a laser stripe image of a track roller; obtaining a target area of ​​a track roller contour in the laser stripe image; performing straight line detection on the target area, marking each edge detection point determined in the straight line detection process; determining the point slope of each edge detection point; obtaining the trend complexity of each pixel point in the target area; determining the laser trend value of each pixel point in the target area; performing connected domain analysis on the target area, obtaining a laser stripe area in the target area; determining each stripe pixel point; determining the corresponding rim thickness of the track roller in the target area based on the curvature of each stripe pixel point, and obtaining the real rim thickness of the track roller in combination with the position of the laser stripe image acquisition device. Thus, the measurement accuracy of the rim thickness of the track roller is improved.

Description

Technical Field

[0001] The present application relates to the technical field of wheel rim thickness measurement, and in particular to an intelligent wheel rim size measurement method and system for supporting wheels. Background Art

[0002] Track rollers are a core load-bearing component in engineering machinery, mainly used to disperse and transfer the weight of the equipment, ensure the smooth operation of the machinery and provide the necessary support. At present, the intelligent measurement of the thickness of the track roller rim generally adopts the detection method of laser as the auxiliary line. By projecting two parallel and equidistant lasers on the track roller, the image of the laser on the track roller is captured, and the thickness of the track roller is calculated by the length of the laser on the track roller and the spacing of the laser on the track roller.

[0003] Since lasers are easily scattered on the metal surface of the roller, the laser detection image is prone to edge blur. For the blurred laser edges, the traditional method uses morphological operations to obtain laser stripe lines and complete thickness detection through laser stripe lines. However, after morphological processing of the laser stripe lines, the laser stripe lines are prone to deformation, resulting in the laser stripe lines being unable to represent the actual laser center lines, resulting in reduced accuracy in measuring the rim size of the roller. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide an intelligent measurement method and system for the rim size of a track roller. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides an intelligent measurement method for the rim size of a track roller, the method comprising the following steps:

[0006] Acquire a laser stripe image of the roller; obtain a target area of ​​the roller contour in the laser stripe image;

[0007] Perform straight line detection on the target area, mark each edge detection point determined in the straight line detection process; determine the point slope of each edge detection point based on the distribution of the slopes of all straight lines intersecting each edge detection point; analyze the discrete degree of the point slopes of a preset number of edge detection points closest to each pixel point in the target area, and determine the trend complexity of each pixel point in the target area;

[0008] Taking the point slope of each edge detection point as a weight, analyzing the distance relationship between each pixel point and all edge detection points in the target area, and combining the trend complexity, determining the laser trend value of each pixel point in the target area; performing a connected domain analysis on the target area to obtain the laser stripe area in the target area; analyzing the difference between the average level of the laser trend values ​​of all pixel points at the same horizontal coordinates in the laser stripe area and the laser trend values ​​of the adjacent pixel points, combining the vertical coordinates of the adjacent pixel points, and screening the stripe pixel points at the same horizontal coordinates;

[0009] Based on the curvature of each stripe pixel point, the corresponding rim thickness of the track roller in the target area is determined, and combined with the position of the laser stripe image acquisition device, the actual rim thickness of the track roller is obtained.

[0010] In one embodiment, the point slope is the average of the slopes of all straight lines intersecting each edge detection point.

[0011] In one embodiment, the trend complexity is the sum of the normalized value of the discreteness and the value 1.

[0012] In one embodiment, the laser trend value is calculated as follows:

[0013] ; In the formula, is the laser trend value of pixel h, is the point slope of edge detection point n, is the trend complexity of pixel h, dis() is the distance calculation function, is a preset value greater than 0, norm() is the normalization function, and N is the number of edge detection points in the target area.

[0014] In one embodiment, the laser stripe region is the first preset number of connected domains in the target region after the connected domains are arranged in descending order according to the areas of the connected domains.

[0015] In one embodiment, the process of determining each stripe pixel point is as follows:

[0016] In any laser stripe region, the reciprocal of the sum of the difference and a preset value greater than 0 is calculated, and the normalized value of the reciprocal is combined with the ordinate of each adjacent pixel point to determine each stripe pixel point.

[0017] In one embodiment, the product of the normalized value of the inverse and the vertical coordinates of each adjacent pixel point is calculated, and the sum of all the products under the same horizontal coordinates is used as the vertical coordinate corresponding to each horizontal coordinate, and the stripe pixel points are determined according to each horizontal coordinate and its corresponding vertical coordinate.

[0018] In one embodiment, the process of determining the corresponding rim thickness of the track roller in the target area is as follows:

[0019] All stripe pixel points in the first laser stripe region are connected in sequence, and the maximum value of the curvature of all stripe pixel points after the connection is calculated. The line connecting all stripe pixel points after the second maximum value is recorded as the first wheel edge stripe, and the second wheel edge stripe in the second laser stripe region is obtained by the same acquisition method as the first wheel edge stripe;

[0020] The average of the number of all pixels on the first wheel rim stripe and the number of all pixels on the second wheel rim stripe is taken as the corresponding wheel rim thickness of the supporting wheel in the target area.

[0021] In one embodiment, the actual rim thickness of the track roller is calculated as follows:

[0022] HL is the actual rim thickness of the track roller, A is the corresponding rim thickness of the track roller in the target area, L is the distance between the two laser stripe image acquisition devices, is the mean value of the ordinates of all pixel points in the first wheel edge stripe, is the mean value of the ordinates of all the pixel points in the second wheel edge stripe.

[0023] In a second aspect, an embodiment of the present application further provides an intelligent rim size measurement system for supporting wheels, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned methods when executing the computer program.

[0024] This application has at least the following beneficial effects:

[0025] The present application collects a laser stripe image of a supporting wheel; obtains a target area of ​​the supporting wheel contour in the laser stripe image; performs straight line detection on the target area and marks each edge detection point determined in the straight line detection process; determines the point slope of each edge detection point based on the distribution of the slopes of all straight lines intersecting each edge detection point; the point slope characterizes the trend of the laser; analyzes the discreteness of the point slopes of a preset number of edge detection points that are closest to each pixel in the target area, and determines the trend complexity of each pixel in the target area; the trend complexity reflects the complexity of the local laser trend in the target area, and improves the accuracy of the subsequent determination of the laser trend value of each pixel point; uses the point slope of each edge detection point as a weight, analyzes the distance relationship between each pixel point in the target area and all edge detection points, and determines the laser trend value of each pixel point in the target area in combination with the trend complexity; the laser trend value is quantified by the laser texture of different areas in the target area, so as to more accurately obtain The laser lines corresponding to the rim thickness of the supporting wheel improve the reliability of the rim thickness measurement of the supporting wheel; the connected domain analysis is performed on the target area to obtain the laser stripe area in the target area; the average level of the laser trend values ​​of all pixel points at the same horizontal coordinates in the laser stripe area and the difference between the laser trend values ​​of adjacent pixel points are analyzed, and the vertical coordinates of the adjacent pixel points are combined to screen and obtain each stripe pixel point at the same horizontal coordinate; the noise influence caused by the irregular edge of the laser stripe image and the deformation of the stripes at the connection point position are avoided by determining the stripe pixel points, so that a more accurate laser texture can be obtained; further, based on the curvature of each stripe pixel point, the corresponding rim thickness of the supporting wheel in the target area is determined, and the actual rim thickness of the supporting wheel is obtained in combination with the position of the laser stripe image acquisition device; the problem of deformation of the laser texture connection point caused by the use of traditional morphological laser texture acquisition algorithms is avoided, and the accuracy of the rim thickness measurement of the supporting wheel is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0027] Figure 1 A flowchart of a method for intelligently measuring the wheel rim size of a track roller provided in one embodiment of the present application;

[0028] Figure 2 Schematic diagram of laser stripe image;

[0029] Figure 3It is a schematic diagram of the target area of ​​the roller profile;

[0030] Figure 4 is a schematic diagram of a line laser;

[0031] Figure 5 Construct a flow chart for the roller flange thickness measurement metric. DETAILED DESCRIPTION

[0032] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of an intelligent measurement method and system for the rim size of a supporting wheel proposed in the present application, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0033] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0034] The specific scheme of an intelligent measurement method and system for the rim size of a track roller provided by the present application is described in detail below with reference to the accompanying drawings.

[0035] See also Figure 1 , which shows a flowchart of a method for intelligently measuring the wheel rim size of a track roller provided by an embodiment of the present application, the method comprising the following steps:

[0036] S01, collecting a laser stripe image of a track roller; obtaining a target area of ​​a track roller contour in the laser stripe image.

[0037] First, install the line laser device and use the camera to take fixed-point photos, including:

[0038] Fix the track roller at the detection position, and install two line laser emitters on a horizontal plane parallel to the ground just above the track roller. The line laser emitted by the line laser emitters is perpendicular to the ground, and the line direction of the laser is parallel to the central axis of the track roller;

[0039] The distance between the emission ports of the two line laser emitters is recorded as the laser distance L;

[0040] On the plane where the line laser emitter is located, a CMOS camera is fixed at the midpoint of the line connecting the two laser emitters, the scene light is turned off to form a dark room, the line laser emitter is turned on, and the captured image is obtained through the CMOS camera;

[0041] The captured image is an image in RGB color space. The captured image is calculated using the grayscale average algorithm and output as a grayscale captured image, which is recorded as a laser stripe image. The schematic diagram of the laser stripe image is shown in Figure 2 As shown; the grayscale average algorithm is a well-known technology. The implementer can select other feasible grayscale conversion algorithms according to the actual situation. This embodiment will not be described in detail here.

[0042] Figure 2 The two bright lines on the middle right are the lasers shining on the rims of the track rollers. The Z-shaped bend in the middle is caused by the protruding rims of the track rollers.

[0043] A region of interest (ROI) is manually set in the laser stripe image. The ROI includes the entire area of ​​the roller contour and is recorded as the target area of ​​the roller contour. The implementer may use other existing algorithms to obtain the target area of ​​the roller contour, which is not limited in this embodiment. The schematic diagram of the target area of ​​the roller contour is shown in FIG. Figure 3 shown.

[0044] exist Figure 3 In the embodiment, since the positions of the camera, the supporting roller and the line laser transmitter are fixed, the laser stripe image is intercepted into a fixed rectangular range to obtain the supporting roller contour target area, which is convenient for subsequent further image processing.

[0045] Due to the scattering of laser on the surface of the supporting wheel and the system noise of the CMOS camera, there is some noise in the laser stripe image, which will interfere with the final measurement. In this embodiment, the median filter method is used to remove the noise; the laser stripe image is a grayscale image. In order to facilitate further processing, the grayscale image needs to be converted into a binary image. Specifically, the laser stripe image is used as input and the median filter algorithm is used for calculation. In this embodiment, the window size is , the output is the filtered laser stripe image; taking the filtered laser stripe image as input, the image binary segmentation algorithm based on Otsu threshold is adopted, and the output is a binary image. In the binary image, the pixel value of 1 represents the laser pixel point, and the pixel value of 0 represents the background pixel point.

[0046] It should be noted that the median filtering algorithm and the image binary segmentation algorithm based on Otsu threshold are both existing well-known technologies. The implementer can select other existing feasible filtering algorithms and image segmentation algorithms according to actual conditions, and this embodiment does not limit them here.

[0047] S02, performing straight line detection on the target area, marking each edge detection point determined in the straight line detection process; determining the point slope of each edge detection point based on the distribution of the slopes of all straight lines intersecting with each edge detection point; analyzing the degree of discreteness of the point slopes of a preset number of edge detection points closest to each pixel point in the target area, and determining the trend complexity of each pixel point in the target area.

[0048] In the binary image, the line laser diagram is as follows Figure 4 As shown, S1, S2, and S3 represent the first laser, and S4, S5, and S6 represent the second laser. Parts of S3 and S6 are irradiated on the rim of the track wheel, parts of S1 and S2 are irradiated on the axle or other parts of the track wheel, and S4 and S5 are irradiated on the connection between the axle or other parts of the track wheel and the rim. Since the diameter of the track wheel rim is larger than other parts of the track wheel, the line laser will appear in the shape shown in Figure 4.

[0049] When measuring the rim thickness by line laser, it is necessary to accurately measure the connection point positions of S2 and S3, and the connection point positions of S5 and S6. Due to the influence of noise, the edge of the line laser is not regular, so the traditional morphological method is used to process the binary image, or the center texture of the binary image is used as the laser texture to complete the rim thickness measurement through the laser texture, which will cause the line laser at the two connection point positions to deform, making it difficult to determine the line laser, affecting the accurate measurement of the rim thickness.

[0050] Therefore, in this embodiment, straight line detection is first performed on the binary image to determine the trend of the line laser, and then the stripe lines are extracted according to the trend to avoid deformation of the connection point position of the image.

[0051] With a binary image as input, the Hough line detection algorithm is used, and the output is multiple straight lines on the binary image. In the Hough line detection process, each edge detection point in the binary image can be determined. The number of edge detection points involved in the Hough line detection algorithm is recorded as N, where the nth edge detection point involves A straight line, The slope of the straight lines is averaged and recorded as the point slope of the nth edge detection point .

[0052] It should be understood that the Hough line detection algorithm and the edge detection points determined in the calculation process thereof are both existing well-known technologies, and will not be described in detail in this embodiment.

[0053] For the nth edge detection point, its point slope It represents the laser trend at its location. Therefore, for any pixel h in the binary image, the point slope of the edge detection point near it can be used to characterize the laser trend at the pixel h.

[0054] like Figure 2 As shown, the line laser will appear as a straight line in some positions and turn in other positions. The trend of the line laser is more complicated at the turning point. For the position where the line laser has a complex trend, the point slope of the edge detection point close to the pixel point at that position should be selected to characterize the laser trend of the pixel point at that position, so as to characterize the variable laser trend at the local position.

[0055] Taking the pixel h in the binary image as an example, obtain the R edge detection points with the closest Euclidean distance to it. In this embodiment, R=30, and the implementer can set it according to the actual situation. This embodiment does not limit it here. The discrete degree of the point slope of the R edge detection points is calculated, and the sum of the normalized value of the discrete degree and the value 1 is used as the trend complexity of the pixel h.

[0056] It should be noted that the degree of dispersion can be calculated by variance, standard deviation, coefficient of variation, etc. This embodiment uses standard deviation as the method for calculating the degree of dispersion, and the normalized value of the degree of dispersion is obtained using the Sigmoid function. The implementer can choose other normalization methods at will, and this embodiment does not limit this.

[0057] The trend complexity ranges from 1 to 2. The larger the value, the more complex the laser trend of the pixel h is. The edge detection point close to the pixel h should be selected to characterize the laser trend of the pixel h.

[0058] S03, taking the point slope of each edge detection point as a weight, analyzing the distance relationship between each pixel point and all edge detection points in the target area, and determining the laser trend value of each pixel point in the target area in combination with the trend complexity; performing a connected domain analysis on the target area to obtain the laser stripe area in the target area; analyzing the difference between the average level of the laser trend values ​​of all pixel points at the same horizontal coordinates in the laser stripe area and the laser trend values ​​of the adjacent pixel points, and combining the vertical coordinates of the adjacent pixel points to screen and obtain each stripe pixel point at the same horizontal coordinates.

[0059] In this embodiment, the laser trend value of pixel point h is calculated as follows:

[0060] ; In the formula, is the laser trend value of pixel h, is the point slope of edge detection point n, is the trend complexity of pixel h, dis() is the distance calculation function, To preset a value greater than 0 to avoid the denominator being 0, norm() is the normalization function, and N is the number of edge detection points in the target area.

[0061] It should be noted that the distance calculation function in this embodiment adopts the Euclidean distance. The implementer may select other distance calculation methods according to the actual situation. The implementer can set it according to the actual situation, and this embodiment does not limit it here.

[0062] It should be understood that in the calculation process of the laser trend value of pixel point h, the trend complexity is used as an adjustment coefficient to control the distance measurement between pixel point h and edge detection point n. The greater the trend complexity, the more complex the laser trend at point h, and the greater the laser trend value.

[0063] Since morphological operations may cause deformation of laser stripes and affect the determination of connection points, the present embodiment extracts laser stripes according to the trend of the line laser.

[0064] like Figure 3 As shown, the trend of the laser stripes in the binary image is from left to right along the horizontal axis of the image, so each horizontal coordinate value in the binary image corresponds to a stripe pixel point.

[0065] The binary image is used as input, and the connected domain detection algorithm is used to calculate, and the output is G connected domains, and the area of ​​each connected domain is output. The G connected domains are arranged in order from large to small in area, and the first two connected domains after arrangement are used as the connected domains of the laser; that is, the connected domain with the largest area in the binary image. The first two connected domains are respectively recorded as the first connected domain and the second connected domain. The first connected domain corresponds to Figure 3 The first laser in the second connected domain corresponds to Figure 3 The second laser in the connected domain. The connected domain detection algorithm is a well-known technology, and this embodiment will not be described in detail. The first connected domain and the second connected domain are both recorded as laser stripe regions.

[0066] This embodiment takes the horizontal coordinate x of the pixel point in the first connected domain as an example to determine the stripe pixel point corresponding to the horizontal coordinate x in the first connected domain. It should be understood that the horizontal coordinates analyzed in this embodiment are all horizontal coordinates corresponding to the pixel points, that is, discrete horizontal coordinate values. The distance between the horizontal coordinate of each pixel point in the first connected domain and the horizontal coordinate x is calculated, and they are arranged from large to small. The first I pixel points after the arrangement are used as the neighboring pixel points of the horizontal coordinate x. In this embodiment, I=30, which can be set by the implementer according to the actual situation. This embodiment does not limit it here. The calculation method of the vertical coordinate value corresponding to the stripe pixel point when the horizontal coordinate is x is:

[0067] ; In the formula, is the ordinate value of the stripe pixel point when the abscissa is x, is the mean value of the laser trend value of all pixels with abscissa x in the first connected domain, is the laser trend value of the neighboring pixel point i of the horizontal coordinate x, norm() is the normalization function, is the ordinate of the neighboring pixel point i of the horizontal coordinate x, I is the number of neighboring pixel points of the horizontal coordinate x, To preset a value greater than 0 to avoid the denominator being 0, in this embodiment The implementer can set it according to the actual situation, and this embodiment does not limit it here.

[0068] The weighted sum of the ordinates of adjacent pixels is used to obtain the ordinate of the stripe pixel when the abscissa is x, thus avoiding the influence of noise caused by the irregular edges of the laser stripe image; combined with the laser trend value, the deformation of the stripes at the connection point is avoided, and the laser texture can be obtained more accurately.

[0069] The method of calculating the ordinate value corresponding to the stripe pixel point when the abscissa is x is used to obtain the stripe pixel points corresponding to each abscissa position in the first connected domain, and all the stripe pixel points in the first connected domain are connected in sequence to form a line, which is recorded as the first laser texture.

[0070] For the second connected domain, the same acquisition method as the first laser texture is adopted to obtain the second laser texture.

[0071] S04, based on the curvature of each stripe pixel point, determine the corresponding rim thickness of the supporting wheel in the target area, and obtain the actual rim thickness of the supporting wheel in combination with the position of the laser stripe image acquisition device.

[0072] After obtaining two laser lines, Figure 4 As shown, to obtain the connection point between line segments S2 and S3, and the connection point between line segments S5 and S6, and then determine the two line segments S3 and S6 through the connection points, line segments S3 and S6 are the laser patterns on the rim of the supporting wheel, which contain the thickness information of the rim. The specific method is:

[0073] Taking the first laser pattern as an example, the number of pixels in the first laser pattern is recorded as K. The curvature calculation method is used to calculate the curvature of each pixel in the first laser pattern. The larger the curvature, the more curved the corresponding position is, and the more likely it is a connection point in the image. The curvature calculation method is an existing known calculation.

[0074] The curvatures of K pixels are arranged from left to right into a sequence, and the sequence is input and filtered using a mean filtering algorithm. In this embodiment, the window length is 3, and the implementer can set it according to actual conditions. The output of the mean filtering algorithm is a filtered sequence; a maximum point detection algorithm is used for the filtered sequence to obtain the maximum point of the filtered sequence. Since a laser texture has two connection points, the maximum point detection algorithm outputs two maximum points, where the pixel point corresponding to the second maximum point is the connection point of the first laser texture; the mean filtering algorithm and the maximum point detection algorithm are both existing publicly known technologies.

[0075] Mark the connection point in the first laser pattern. Figure 4 corresponds to the connecting point of line segments S2 and S3; all the stripe pixels on the right side of the connecting point, including the connecting point, form a stripe, recorded as the first wheel edge stripe, corresponding to Figure 4 Line segment S3 in .

[0076] The second laser pattern is processed in the same way as the first wheel rim pattern, and the second wheel rim pattern is obtained, which corresponds to Figure 4 Line segment S6 in .

[0077] The first rim stripe and the second rim stripe contain information about the rim.

[0078] First, according to the first and second wheel rim stripes, calculate the corresponding wheel rim thickness A in the target area. The calculation method is:

[0079] The number of pixels of the first wheel edge stripe is recorded as , the number of pixels of the second wheel edge stripe is recorded as , then the corresponding rim thickness A of the track roller in the target area is and The mean of .

[0080] The actual wheel flange thickness of the supporting wheel is further calculated based on the first wheel flange stripe and the second wheel flange stripe. The specific method is as follows:

[0081] HL is the actual rim thickness of the track roller, A is the corresponding rim thickness of the track roller in the target area, L is the distance between the two laser stripe image acquisition devices, is the mean value of the ordinates of all pixel points in the first wheel edge stripe, is the average value of the vertical coordinates of all pixel points in the second wheel rim stripe. The flow chart of the construction of the roller rim thickness measurement index is as follows: Figure 5 shown.

[0082] At this point, the measurement of the rim thickness of the track roller can be completed.

[0083] Based on the same inventive concept as the above method, an embodiment of the present application also provides an intelligent measurement system for the rim size of a supporting wheel, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above-mentioned intelligent measurement methods for the rim size of a supporting wheel are implemented.

[0084] It should be noted that the above sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0085] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0086] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present application should be included in the protection scope of the present application.

Claims

1. An intelligent measurement method for the wheel rim size of a track roller, characterized in that: The method comprises the following steps: Acquire a laser stripe image of the roller; obtain a target area of ​​the roller contour in the laser stripe image; Perform straight line detection on the target area, mark each edge detection point determined in the straight line detection process; determine the point slope of each edge detection point based on the distribution of the slopes of all straight lines intersecting each edge detection point; analyze the discrete degree of the point slopes of a preset number of edge detection points closest to each pixel point in the target area, and determine the trend complexity of each pixel point in the target area; Taking the point slope of each edge detection point as a weight, analyzing the distance relationship between each pixel point and all edge detection points in the target area, and combining the trend complexity, determining the laser trend value of each pixel point in the target area; performing a connected domain analysis on the target area to obtain the laser stripe area in the target area; analyzing the difference between the average level of the laser trend values ​​of all pixel points at the same horizontal coordinates in the laser stripe area and the laser trend values ​​of the adjacent pixel points, combining the vertical coordinates of the adjacent pixel points, and screening the stripe pixel points at the same horizontal coordinates; Based on the curvature of each stripe pixel point, the corresponding rim thickness of the track roller in the target area is determined, and combined with the position of the laser stripe image acquisition device, the actual rim thickness of the track roller is obtained; The process of determining each stripe pixel point is as follows: In any laser stripe region, the inverse of the sum of the difference and a preset value greater than 0 is calculated, and the normalized value of the inverse is combined with the ordinate of each adjacent pixel point to determine each stripe pixel point; The process of determining the corresponding wheel rim thickness of the track roller in the target area is as follows: All stripe pixel points in the first laser stripe region are connected in sequence, and the maximum value of the curvature of all stripe pixel points after the connection is calculated. The line connecting all stripe pixel points after the second maximum value is recorded as the first wheel edge stripe, and the second wheel edge stripe in the second laser stripe region is obtained by the same acquisition method as the first wheel edge stripe; The average of the number of all pixels on the first wheel rim stripe and the number of all pixels on the second wheel rim stripe is taken as the corresponding wheel rim thickness of the supporting wheel in the target area; The actual flange thickness of the track roller is calculated as follows: HL is the actual rim thickness of the track roller, A is the corresponding rim thickness of the track roller in the target area, L is the distance between the two laser stripe image acquisition devices, is the mean value of the ordinates of all pixel points in the first wheel edge stripe, is the mean value of the ordinates of all the pixel points in the second wheel edge stripe.

2. The intelligent measurement method for the wheel rim size of a track wheel according to claim 1, characterized in that: The point slope is the average of the slopes of all straight lines intersecting each edge detection point.

3. The intelligent measurement method for the wheel rim size of a track wheel according to claim 1, characterized in that: The trend complexity is the sum of the normalized value of the discreteness and the value 1.

4. The intelligent measurement method for the wheel rim size of a track wheel according to claim 1, characterized in that: The calculation method of the laser trend value is: ; In the formula, is the laser trend value of pixel h, is the point slope of edge detection point n, is the trend complexity of pixel h, dis() is the distance calculation function, is a preset value greater than 0, norm() is the normalization function, and N is the number of edge detection points in the target area.

5. The intelligent measurement method for the wheel rim size of a track wheel according to claim 1, characterized in that: The laser stripe region is the first preset number of connected domains in the target region after the connected domains are arranged in descending order according to the areas of the connected domains.

6. The intelligent measurement method for the wheel rim size of a track wheel according to claim 1, characterized in that: The product of the normalized value of the reciprocal and the ordinate of each adjacent pixel point is calculated, the sum of all the products under the same abscissa is used as the ordinate corresponding to each abscissa, and each stripe pixel point is determined according to each abscissa and its corresponding ordinate.

7. An intelligent measurement system for the rim size of a track wheel, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

Images