A method for measuring the particle size of runway long-wave unevenness resonance fracture.

Abstract

This application relates to the field of airport runway inspection and maintenance technology, specifically to a method for measuring the fragmentation particle size of runway long-wave irregularity resonant fracture. The method includes: evaluating the irregularity based on three-dimensional elevation data of the airport runway; acquiring a grayscale image of the fractured runway after resonant fracture; obtaining edge information, identifying suspected edge discontinuity points through edge endpoints and edge trend directions, and performing gradient reconstruction processing on the fractured grayscale image using grayscale boundary balance values ​​and gradient adjustment coefficients to obtain a gradient reconstruction map; achieving accurate identification and enhancement of edge discontinuity points, improving the image's ability to identify discontinuous edges, while avoiding misjudging internal textures as aggregate particle boundaries; and using a watershed algorithm based on control markers to segment aggregate particles and measure the fractured particle size, achieving comprehensive identification and quantitative analysis of runway fracture particle size, and guiding resonant fracture construction.

Description

Technical Field

[0001] This application relates to the field of airport runway inspection and maintenance technology, specifically to a method for measuring the particle size of runway long-wave unevenness resonance fracture. Background Technology

[0002] Runway unevenness can cause severe bouncing, pitching, and vibration during aircraft taxiing, increasing handling difficulty and accelerating damage to aircraft components and the runway surface. During high-speed taxiing, long-wave unevenness of the runway surface significantly affects aircraft vibration acceleration, and runway damage further deteriorates runway smoothness. For runways with poor unevenness assessments, traditional airport runway maintenance methods cannot fundamentally address the overall structural degradation, necessitating comprehensive resonant fracture reconstruction. Measuring and controlling the runway resonant fracture particle size helps reduce the risk of reflective cracking after reconstruction and enhances the load-bearing capacity and durability of in-situ regeneration.

[0003] Currently, the main method for measuring crushed particle size is image processing. Image processing uses a high-definition industrial camera to capture surface images of the crushed concrete layer of the runway and combines this with GNSS positioning data for comprehensive measurement. However, this method has difficulties in segmenting aggregate particle boundaries, especially when particles are stacked or adhered at the edges, easily leading to under-segmentation and significant measurement errors in crushed particle size. Summary of the Invention

[0004] In view of the above, it is necessary to provide a method for measuring the particle size of runway long-wave unevenness resonance fracture to solve the above problems.

[0005] One embodiment of this application provides a method for measuring the particle size of runway long-wave irregularity resonance fracture, the method comprising:

[0006] Based on the three-dimensional elevation model of the pre-set test section of the airport runway, the unevenness of the test section of the airport runway is evaluated, the road surface reconstruction area is obtained and resonant fracturing is performed, and fracturing grayscale images are collected.

[0007] Edge detection is performed on the fragmented grayscale image. Based on the distribution characteristics of edge points, the edge endpoints are determined. Each edge endpoint is encoded as a starting point. Combining the differences between encoding orders, gradient direction, and gradient magnitude distribution, the terminal gradient direction angle and the nearest point of each edge endpoint are obtained. Based on the distribution characteristics of non-edge points along the line connecting the nearest point to the corresponding edge endpoint, the continuation line of each edge endpoint is obtained, and suspected discontinuous edges are screened out. The gradient magnitude difference between each suspected discontinuous point on each suspected discontinuous edge and all non-edge points in the fragmented grayscale image is compared. Combined with the terminal gradient direction angle, the edge discontinuity confidence of each suspected discontinuous point is determined.

[0008] Based on the difference in grayscale dispersion of pixels in different parts of the preset local window of each pixel, the grayscale boundary balance value of each pixel is determined. Combined with the edge discontinuity confidence of each suspected discontinuity point, the fragmented grayscale image is reconstructed by gradient to obtain the gradient reconstruction map.

[0009] The gradient reconstruction map is segmented to obtain the boundary lines of the aggregate particles, and the crushed particle size of the aggregate particles is obtained.

[0010] Preferably, the road surface reconstruction area is specifically the airport runway with the worst unevenness evaluation value.

[0011] Preferably, the specific process for determining the edge endpoint is as follows:

[0012] The number of edge points is counted by counting the eight neighboring pixels of each edge point. Edge points with a count of 1 are recorded as edge termination points.

[0013] Preferably, obtaining the terminal gradient direction angle and the nearest point of each edge endpoint specifically involves:

[0014] The edge binary image of the broken grayscale image is encoded using the chain code method, and the difference in the order of occurrence of two edge points in the chain code is taken as the chain code distance.

[0015] Edge points whose chain code distance from the edge endpoint is less than or equal to a preset value are considered as neighboring points of the edge endpoint; the average of the angles between the gradient directions of all neighboring points of the edge endpoint and the horizontal direction is taken as the endpoint gradient direction angle of the edge endpoint.

[0016] The nearest point to the edge end point is the one that is 1 distance from the chain code of the edge end point and has the largest gradient magnitude.

[0017] Preferably, obtaining the continuation line of each edge endpoint and filtering out suspected discontinuous edges specifically involves:

[0018] Starting from the edge endpoint, search for a continuation growth point along the connecting line; if the continuation growth point is not an edge point, continue searching for the next point along the connecting line; if the continuation growth point is an edge point, growth stops; the line connecting all the continuation growth points is recorded as the continuation line of the edge endpoint.

[0019] The number of all continuous growth points on the continuous line is taken as the edge continuity length, and the continuous line with an edge continuity length greater than the preset length is taken as a suspected discontinuous edge.

[0020] Preferably, the process of determining the edge discontinuity confidence level of each suspected discontinuity point is as follows:

[0021] Calculate the difference between the gradient magnitude of each suspected discontinuous point of each suspected discontinuous edge and the average gradient magnitude of all non-edge points in the fragmented grayscale image;

[0022] Calculate the average angle between the gradient direction and the horizontal direction of all suspected discontinuous points in each suspected discontinuous edge to obtain the discontinuous gradient direction angle;

[0023] Obtain the cosine of the angle between the standard vector corresponding to the end gradient direction angle of each suspected discontinuous edge and the standard vector corresponding to the discontinuous gradient direction angle;

[0024] The difference between each suspected discontinuity point and the cosine value of the angle between them and the suspected discontinuity edge is positively fused to obtain the edge discontinuity confidence of each suspected discontinuity point on each suspected discontinuity edge.

[0025] Preferably, the process of determining the grayscale boundary balance value of each pixel is as follows:

[0026] The gradient direction of each pixel is obtained as the tangent line at each pixel. The local window of each pixel is segmented using the tangent line, and the standard deviation of gray values ​​of all pixels in each segment is obtained. The negative correlation mapping result of the difference between all gray standard deviations obtained from the local window of each pixel is used as the gray value boundary balance value of each pixel.

[0027] Preferably, the specific process of gradient reconstruction of the fragmented grayscale image is as follows:

[0028] In a fragmented grayscale image, if a pixel is a suspected discontinuity point, the gradient adjustment coefficient of the suspected discontinuity point is obtained based on the grayscale boundary balance value of the suspected discontinuity point and the edge discontinuity confidence. Through calculation The result is used to obtain the gradient magnitude after adjustment for suspected discontinuity points; where, This represents the gradient magnitude before pixel adjustment. This represents the preset first threshold. This represents the average gradient magnitude of all pixels within a preset local window of a pixel.

[0029] If a pixel is an edge point, the normalized value of the grayscale boundary balance of the edge point is used as the gradient reduction coefficient of the edge point. When the gradient reduction coefficient is less than the preset second threshold At that time, through calculation The result is used to obtain the adjusted gradient magnitude of the corresponding edge point; otherwise, the gradient magnitude of the edge point remains unchanged.

[0030] The gradient magnitude of the remaining pixels remains unchanged.

[0031] Preferably, the gradient adjustment coefficient is a normalized value of the ratio of the edge discontinuity confidence of the suspected discontinuity point to the gray-level boundary balance value.

[0032] Preferably, the crushed particle size of the aggregate particles is determined by the length of the minimum bounding rectangle of the base material particle boundary.

[0033] This application has at least the following beneficial effects:

[0034] First, this application uses a three-dimensional elevation model to evaluate the unevenness of the airport runway, which can identify areas that need repair. By combining a vehicle-mounted laser profiler with a GNSS system, it can quickly acquire three-dimensional elevation data and unevenness evaluation of the runway, avoiding the problem of low accuracy of vehicle-mounted laser profilers in identifying long-wavelength unevenness of airport runways. At the same time, the high-definition industrial camera is linked with the GNSS system to achieve comprehensive image acquisition of the runway fracture layer, with high measurement efficiency and wide coverage.

[0035] Furthermore, edge detection is performed on the fragmented grayscale image to determine the location of edge endpoints and provide basic data for subsequent analysis. Edge points are encoded and analyzed in conjunction with gradient direction and magnitude information to obtain the terminal gradient direction angle and nearest point of each edge endpoint. This helps determine the characteristics of each edge endpoint and facilitates a more accurate understanding of edge morphology and direction. By analyzing the distribution characteristics of non-edge points along the line connecting the nearest point to the corresponding edge endpoint, possible discontinuous edges are screened out. This effectively distinguishes between continuous edges and broken sections, aiding in a refined analysis of runway surface damage. By analyzing the difference between the gradient magnitude of discontinuous points and that of non-edge points, combined with the terminal gradient direction angle... Evaluating the confidence level of each suspected discontinuity point helps determine which points are genuine edge discontinuities and which are errors or noise. By analyzing the grayscale boundary value of each pixel and combining it with the discontinuity confidence level of suspected discontinuities, the accuracy of discontinuous edges can be improved. Gradient reconstruction combined with grayscale boundary balance values ​​enhances the gradient of edge discontinuities, improving the accuracy of the watershed segmentation algorithm in identifying watershed ridges, avoiding missegmentation of aggregate particles due to edge discontinuity, and enhancing the segmentation ability between adjacent aggregate particles. Gradient reduction is applied to edge points of aggregate particles that belong to internal textures to avoid misjudging internal textures as aggregate particle boundaries and to prevent defects such as internal pores of aggregate particles from being misjudged as aggregate particles, thus preventing increased measurement error in crushed particle size.

[0036] Finally, by segmenting aggregate particles using a watershed algorithm based on control markers, the accuracy of the watershed algorithm in identifying aggregate particle boundaries is improved, as is the accuracy of crushed particle size measurement. This enables comprehensive identification and quantitative analysis of runway crushed particle size, guiding the construction of frequency-controlled crushing and runway repair. Attached Figure Description

[0037] Figure 1 A flowchart of a method for measuring the fragmentation particle size of long-wave uneven resonance fracture of a runway, provided in this application;

[0038] Figure 2 A flowchart for obtaining the edge discontinuity confidence level provided in this application. Detailed Implementation

[0039] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.

[0040] Unless otherwise defined, 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. The terminology used in this application's specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0041] It should also be noted that the terms "first" and "second" in this application and its accompanying drawings are used to distinguish similar objects, rather than to describe a specific order or sequence. The methods disclosed in the embodiments of this application or the methods shown in the flowcharts include one or more steps for implementing the method. Without departing from the scope of protection of this application, the execution order of multiple steps can be interchanged, and some steps can also be deleted.

[0042] Unless otherwise defined, 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 pertains.

[0043] This application proposes a method for measuring the particle size of runway long-wave unevenness resonance fracture, which is applied in the field of airport runway inspection and maintenance technology. (See attached document.) Figure 1 The method includes the following steps:

[0044] The first step: Based on the three-dimensional elevation model of the pre-set test section of the airport runway, evaluate the unevenness of the test section of the airport runway, obtain the road surface reconstruction area and perform resonance fracturing, and collect the fracturing grayscale image.

[0045] Since laser profilers cannot accurately acquire long-wavelength data of airport runways, this application combines a vehicle-mounted laser profiler with a Global Navigation Satellite System (GNSS) to construct an airport runway testing system. The sampling interval of the vehicle-mounted laser profiler is 0.025m, and the sampling frequency of the GNSS system is 1Hz. Starting from the runway start point, it is ensured that the vehicle-mounted laser profiler and the GNSS system receive data synchronously. Specifically, the vehicle-mounted laser profiler includes a laser displacement sensor, an acceleration sensor, and a distance sensor. The GNSS system transmits data in network mode, and high-precision positioning is obtained using real-time differential technology and CORS differential correction data.

[0046] This application aims to obtain automatic three-dimensional elevation data of airport runways. The runway width centerline and a 10m distance from the centerline are designated as the test interval. 21 test paths are obtained at 1m intervals along the width of the test interval. The airport runway testing system is used to conduct full-band unevenness tests on all test paths to obtain the three-dimensional elevation data of all test paths.

[0047] During high-speed taxiing, long-wave irregularities on the runway surface significantly affect the aircraft's vibration acceleration. This application obtains a three-dimensional elevation model of the airport runway within the test section based on the three-dimensional elevation data of all test paths. The acquisition of the three-dimensional elevation model is a well-known existing technique, and this application will not elaborate on it.

[0048] According to the airport runway smoothness evaluation method disclosed in the literature "Airport runway smoothness evaluation method based on time-frequency analysis", the airport runway elevation time-frequency analysis is carried out based on the three-dimensional elevation model of the airport runway on the test section, and the unevenness evaluation value at any position on the airport runway in the test section is obtained through the runway smoothness evaluation surface.

[0049] When the unevenness rating is Level 1 "Good", the airport runway has high smoothness, the aircraft vibration response is small, and the aircraft operation is safe. When the unevenness rating is Level 2 "Medium", the airport runway has poor smoothness, which will seriously interfere with the instruments in the cockpit and cause discomfort to the aircraft passengers. It is not necessary to close the affected runway, but the runway needs to be repaired in time. When the unevenness rating is Level 3 "Poor", the airport runway has extremely poor smoothness. The long-wave unevenness of the runway area increases the safety risk of aircraft taxiing at high speeds. The affected runway should be closed immediately and repaired using the overall road surface fragmentation and reconstruction technology.

[0050] The airport runway with an unevenness rating of "poor" (third level) was selected as the pavement reconstruction area. A pavement resonance crusher was used, which generates a vibration frequency of 42-45Hz, consistent with the natural frequency of the cement concrete pavement, thus creating a resonance effect. This causes the old concrete slab to break from top to bottom, exhibiting a gradual change in depth from "fine at the top to coarse at the bottom." The interlocking and interlocking structure between the aggregate particles after resonance crushing provides better load transfer.

[0051] This application employs a high-definition industrial camera to acquire images of the compacted fractured concrete layer of the airport runway. The camera lens is pointed vertically at the ground, 0.5m above the ground, with a shooting range of a 0.5m × 0.5m rectangular area. RGB images of the crushed aggregate particles are acquired, and the shooting position is recorded in real time using a GNSS system to ensure complete coverage of the fractured concrete layer. The RGB images of the crushed aggregate particles are then converted to grayscale to obtain a fractured grayscale image with 256 gray levels.

[0052] The second step is to perform edge detection on the fragmented grayscale image, determine the edge endpoints based on the distribution characteristics of the edge points, encode each edge point starting from an edge endpoint, and obtain the endpoint gradient direction angle and nearest point for each edge endpoint by combining the differences between encoding orders, gradient direction, and gradient magnitude distribution. Based on the distribution characteristics of non-edge points along the line connecting the nearest point to the corresponding edge endpoint, obtain the continuation line of each edge endpoint and filter out suspected discontinuous edges. Compare the gradient magnitude difference between each suspected discontinuous point on each suspected discontinuous edge and all non-edge points in the fragmented grayscale image, and determine the edge discontinuity confidence of each suspected discontinuous point by combining the endpoint gradient direction angle.

[0053] After resonance crushing, the aggregate particles of the runway concrete are often numerous and densely distributed, and the aggregate particles have different shapes. After the cement concrete slab is crushed, the aggregate particles stack and interlock with each other, which can easily cause edge adhesion. Some aggregate particles have low edge gradients and indistinct boundary features, which can easily cause discontinuous edges.

[0054] This application uses a fragmented grayscale image as input to the Sobel edge detection algorithm, obtaining the gradient magnitude and direction of any pixel in the fragmented grayscale image, as well as the binary edge map of the fragmented grayscale image. It then counts the number of 8-neighbor pixels of any edge point in the binary edge map that are also edge points. ,Will The edge points are used as the edge endpoints of the broken grayscale image.

[0055] The edges of aggregate particles after runway resonance crushing exhibit local similarities, and pixels at discontinuous edge locations are highly likely to match the edge trend characteristics of edge endpoints, with relatively short discontinuity lengths. This application uses edge endpoint x as an example, employing the binary edge image as input to the 8-direction chain code method, with edge endpoint x as the starting point. The difference in the order of occurrence of two edge points in the chain code is used as the chain code distance. For example, if there is an edge point y in the 8-neighborhood of edge endpoint x, with chain code order 1 for edge endpoint x and chain code order 2 for edge point y, then the chain code distance between edge endpoint x and edge point y is 1. All chain code distances not greater than... The edge point, as the neighboring point of the edge endpoint. This represents a preset value, which is 5 in this embodiment. The mean angle between the gradient direction of all neighboring points of the edge endpoint and the horizontal direction in radians is calculated and used as the endpoint gradient direction angle of the edge endpoint x.

[0056] Among all neighboring points, the nearest point to the edge endpoint x with a distance of 1 from the chain code is taken as the closest point to the edge endpoint x. It should be noted that if the number of neighboring points with a distance of 1 from the chain code of the edge endpoint x is greater than 1, all neighboring points with a distance of 1 from the chain code of the edge endpoint x are recorded as the nearest filtered points, and the nearest filtered point with the largest gradient magnitude is taken as the closest point to the edge endpoint x.

[0057] Connect the nearest point and the edge endpoint with a line pointing from the nearest point to the edge endpoint. Record this line direction as the edge trend direction of edge endpoint x. Starting from edge endpoint x, obtain continuation growth points along the edge trend direction for continued growth. If a continuation growth point is not an edge point, continue obtaining the next continuation growth point along the edge trend direction. When a continuation growth point becomes an edge point, stop the continuation growth. Obtain all continuation growth points of edge endpoint x, and record the line connecting all continuation growth points as the continuation line of edge endpoint x.

[0058] Since the discontinuous length of aggregate particle edges is often low, this application counts the total number of all continuous growth points on the continuity line of the edge end point, denoted as the edge continuity length. From the set of continuity lines corresponding to all edge end points, those with an edge continuity length not greater than [a certain value] are eliminated. The continuation lines are then marked as suspected discontinuous edges. This represents the preset length, which is 15 in this embodiment.

[0059] In a fragmented grayscale image, the mean of the gradient magnitudes of all non-edge points is obtained, and this mean is denoted as the non-edge average gradient of the fragmented grayscale image. Calculate the difference between the gradient magnitude of each suspected discontinuous point of each suspected discontinuous edge and the non-edge average gradient of the broken grayscale image.

[0060] All pixels on the suspected discontinuous edge are recorded as suspected discontinuous points; the average of the angle between the gradient direction of all suspected discontinuous points in each suspected discontinuous edge and the horizontal radian angle is obtained as the discontinuity gradient direction angle of the suspected discontinuous edge.

[0061] Obtain the cosine value of the angle between the standard vector corresponding to the end gradient direction angle of each suspected discontinuous edge and the standard vector corresponding to the discontinuous gradient direction angle. Forwardly fuse the difference of each suspected discontinuous point with the cosine value of the angle obtained from the suspected discontinuous edge to obtain the edge discontinuity confidence of each suspected discontinuous point on each suspected discontinuous edge.

[0062] Specifically, this application obtains the edge discontinuity confidence of the j-th suspected discontinuity point in the i-th suspected discontinuity edge using the following formula. :

[0063]

[0064] In the formula, It is the gradient magnitude of the j-th suspected discontinuity point in the i-th suspected discontinuity edge. It is the non-edge average gradient of a fragmented grayscale image. It is the standard vector corresponding to the terminal gradient direction angle of the i-th suspected discontinuous edge's end point. It is the standard vector corresponding to the discontinuous gradient direction angle of the i-th suspected discontinuous edge. To calculate the cosine of the angle between the standard vector corresponding to the terminal gradient direction angle and the standard vector corresponding to the discontinuous gradient direction angle.

[0065] Used to reflect the gradient difference between suspected discontinuity points and non-edge pixels. The larger the value, the more significant the gradient strength of the suspected discontinuous point compared to the non-edge point, which is consistent with the gradient behavior characteristics of edge discontinuity. Used to reflect the consistency of gradient direction between suspected discontinuous edges and neighboring edges. The larger the value, the more locally similar the suspected discontinuous edge is to its neighboring edges in the gradient direction. This indicates that the suspected discontinuous edge is more likely to be a discontinuous edge on the edge where the edge's endpoint is located. Therefore, a higher confidence weight should be given to the suspected discontinuous points on the suspected discontinuous edge. The larger the value, the better. The flowchart for obtaining the edge discontinuity confidence score is shown below. Figure 2 As shown.

[0066] The third step: Based on the difference in grayscale dispersion of pixels in different parts of the preset local window of each pixel, determine the grayscale boundary balance value of each pixel, and combine the edge discontinuity confidence of each suspected discontinuity point to perform gradient reconstruction of the fragmented grayscale image to obtain the gradient reconstruction map.

[0067] In the grayscale image of the broken aggregate, due to the irregularity of the aggregate particles after resonance crushing, there are some texture features inside the boundary of the aggregate particles, such as crushed stone grooves and irregular protrusions. These textures inside the aggregate particles usually also have strong gradient features and are easily regarded as the edge of the aggregate particles or the edge discontinuity point.

[0068] For each pixel in the fragmented grayscale image, the gradient direction of the pixel is obtained. The straight line perpendicular to the gradient direction of the pixel is taken as the tangent line of the pixel, and the pixel grayscale value changes with the least change on the tangent line. In this application, an n*n local window is constructed with each pixel as the center in the fragmented grayscale image. The n*n local window is used to reflect the local grayscale characteristics of the pixel. In this embodiment, a 7*7 window is used.

[0069] The local window is divided into two parts by using a tangent line passing through its center. The pixels within these two parts are designated as the first local pixel and the second local pixel, respectively. It should be noted that pixels located on the tangent line are not included in the division.

[0070] This application obtains the grayscale boundary balance value of a pixel using the following formula. :

[0071]

[0072] In the formula, It is the standard deviation of the grayscale values ​​of all first local pixels within the local window. It is the standard deviation of the grayscale values ​​of all second local pixels within the local window. It is a parameter tuning factor, which is used to prevent the denominator from being 0. Its value range is set to a positive number not greater than 0.01. In this embodiment, it is set to 0.001.

[0073] This is used to reflect the grayscale balance characteristics of the two parts of pixels obtained by segmentation along the tangent direction. If the pixel is considered as the internal texture of the aggregate particle, then the pixels on both sides of the tangent direction belong to the same aggregate particle, and their grayscale distribution is relatively similar. The smaller the value, the better the grayscale balance value. The larger the value, the greater the difference in grayscale distribution. If a pixel is used as the boundary of an aggregate particle, then pixels on both sides of the tangent direction correspond to different aggregate particles, resulting in significant differences in grayscale distribution. The larger the value, the better the grayscale boundary balance value. The smaller.

[0074] This application uses the normalized value of the ratio of the edge discontinuity confidence level to the gray-level boundary balance value of each suspected discontinuity point as the gradient adjustment coefficient of the suspected discontinuity point, which is used to reflect the strength of the gradient magnitude adjustment required for the suspected discontinuity point.

[0075] The higher the confidence level of edge discontinuity, the higher the probability that the suspected discontinuity point belongs to edge discontinuity. The gradient magnitude of the suspected discontinuity point should be increased, and the larger the gradient adjustment coefficient, the better the ability of the fragmented grayscale image to identify discontinuous edges.

[0076] The purpose of calculating the grayscale boundary balance value is to eliminate the internal texture of aggregate particles that are easily misjudged as discontinuous edge points. The larger the grayscale boundary balance value, the closer the grayscale distribution of the two parts divided by the tangent direction is. The more likely the suspected discontinuous point corresponding to the pixel is the internal texture of the aggregate particle, the weaker the gradient amplitude of the suspected discontinuous point should be. The smaller the gradient adjustment coefficient, the less likely the internal texture in the broken grayscale image is to be misjudged as the boundary of the aggregate particle.

[0077] The edges obtained by the Sobel edge detection algorithm in the broken grayscale image may also be pseudo-boundaries, representing the internal texture of the aggregate particles. The grayscale boundary balance values ​​of all edge points are balanced and normalized. The normalized result is used as the gradient reduction coefficient for the edge points. In this embodiment, the normalization method uses the maximum-minimum normalization method.

[0078] This application performs gradient reconstruction processing on fragmented grayscale images in the following ways to obtain gradient reconstruction maps:

[0079] (1) For the case where pixels in a fragmented grayscale image are suspected discontinuous points: This application obtains the gradient adjustment coefficient of the suspected discontinuous points. Gradient magnitude The mean gradient magnitude of all pixels within a 5x5 local window at the suspected discontinuous point. First threshold Gradient reconstruction is performed using the following formula:

[0080]

[0081] Let be the gradient magnitude after pixel gradient reconstruction. From the formula, we know that if... The higher the probability that a suspected discontinuity point belongs to the discontinuous edge of an aggregate particle, the more likely a gradient enhancement adjustment should be made to the suspected discontinuity point. The higher the probability that a suspected discontinuity point belongs to the internal texture of the aggregate particle, the more gradient weakening adjustment is applied to the suspected discontinuity point. In the formula, the first threshold... In this embodiment, 0.5 is used.

[0082] (2) For pixels in a broken grayscale image that are edge points: Since the gradient of edge points is already high, and to avoid internal textures being misjudged as aggregate particle edges, this application only performs gradient reduction processing on edge points that may be internal textures. Specifically, this application obtains the gradient reduction coefficient of edge points. Gradient magnitude The average gradient magnitude of all pixels within a 5x5 local window at the edge point Second threshold Gradient reconstruction is performed using the following formula:

[0083]

[0084] Let be the gradient magnitude after pixel gradient reconstruction. From the formula, we know that if... The higher the probability that an edge point belongs to the internal texture of an aggregate particle, the more gradient weakening processing is applied to the edge point. In the formula, the second threshold... In this embodiment, the value is 0.6.

[0085] (3) For pixels in a fragmented grayscale image that are neither edge points nor suspected discontinuities, this application obtains the gradient magnitude of the pixels. Gradient reconstruction based on gradient adjustment is performed using the following formula:

[0086]

[0087] In the formula, It is the gradient magnitude after pixel gradient reconstruction.

[0088] The fourth step is to segment the gradient reconstruction map to obtain the boundary lines of the aggregate particles and obtain the crushed particle size of the aggregate particles.

[0089] Because the aggregate particles after resonant crushing are exposed, they have high reflectivity, appear bright, and have high grayscale values, while the gaps between aggregate particles are often dark and have low grayscale values. This application uses the crushed grayscale image as input to the Otsu's Maximum Inter-Class Variance (OSTU) algorithm. Pixels with grayscale values ​​higher than the segmentation threshold are considered foreground and binarized to 1, while other pixels are binarized to 0. The binarization result is recorded as the foreground image. To eliminate isolated foreground pixels in the foreground image, this application performs morphological optimization on the foreground image using an opening reconstruction operation. The foreground image after the opening reconstruction operation is used as an internal control marker. .

[0090] To obtain the external control markers, this application inverts the foreground image to obtain the first background image, and then processes the internal control markers. After performing a dilation operation, a negation operation is performed to obtain a second background image. The first and second background images are then combined to obtain the external control marker. .

[0091] This application obtains the gradient reconstruction image after gradient reconstruction of a fragmented grayscale image. internal control flags External control markers and gradient reconstruction graph As input to the watershed algorithm based on control markers, the watershed ridge line is obtained, each aggregate particle is segmented, and the boundary line of each aggregate particle is obtained. The watershed algorithm based on control markers is a well-known existing technology, and will not be described in detail in this application.

[0092] The boundary line of each aggregate particle is fitted with a minimum bounding rectangle, and the length of the minimum bounding rectangle is taken as the crushed pixel diameter of the aggregate particle. The pixel coordinates of each pixel in the crushed grayscale image are obtained, with a default depth of 0.5m. The pixel coordinates are converted to world coordinates, thus obtaining the crushed particle diameter of each aggregate particle, completing the measurement of the crushed particle diameter. Image technology enables comprehensive identification and quantitative analysis of runway crushed particle diameter, with fast measurement speed. Based on the crushed particle diameter distribution, it facilitates subsequent guidance of frequency-controlled crushing construction, reduces the risk of reflective cracking after reconstruction, and enhances the load-bearing capacity and durability of in-situ regeneration of airport runways.

[0093] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description; sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0094] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for measuring the particle size of broken particles in long-wave irregularity resonance crushing of a runway, characterized in that, The method includes the following steps: Based on the three-dimensional elevation model of the pre-set test section of the airport runway, the unevenness of the test section of the airport runway is evaluated, the road surface reconstruction area is obtained and resonant fracturing is performed, and fracturing grayscale images are collected. Edge detection is performed on the fragmented grayscale image. Based on the distribution characteristics of edge points, the edge endpoints are determined. Each edge endpoint is encoded as a starting point. Combining the differences between encoding orders, gradient direction, and gradient magnitude distribution, the terminal gradient direction angle and the nearest point of each edge endpoint are obtained. Based on the distribution characteristics of non-edge points along the line connecting the nearest point to the corresponding edge endpoint, the continuation line of each edge endpoint is obtained, and suspected discontinuous edges are screened out. The gradient magnitude difference between each suspected discontinuous point on each suspected discontinuous edge and all non-edge points in the fragmented grayscale image is compared. Combined with the terminal gradient direction angle, the edge discontinuity confidence of each suspected discontinuous point is determined. Based on the difference in grayscale dispersion of pixels in different parts of the preset local window of each pixel, the grayscale boundary balance value of each pixel is determined. Combined with the edge discontinuity confidence of each suspected discontinuity point, the fragmented grayscale image is reconstructed by gradient to obtain the gradient reconstruction map. The gradient reconstruction map is segmented to obtain the boundary lines of the aggregate particles, and the crushed particle size of the aggregate particles is obtained. The specific steps for obtaining the terminal gradient direction angle and the nearest point of each edge endpoint are as follows: The edge binary image of the broken grayscale image is encoded using the chain code method, and the difference in the order of occurrence of two edge points in the chain code is taken as the chain code distance. Edge points whose chain code distance from the edge endpoint is less than or equal to a preset value are considered as neighboring points of the edge endpoint; the average of the angles between the gradient directions of all neighboring points of the edge endpoint and the horizontal direction is taken as the endpoint gradient direction angle of the edge endpoint. The nearest point to the edge end point is the one that is 1 distance from the chain code of the edge end point and has the largest gradient magnitude.

2. The method for measuring the particle size of runway long-wave unevenness resonance fracture as described in claim 1, characterized in that, The road surface reconstruction area specifically refers to the airport runway with the worst unevenness evaluation value.

3. The method for measuring the particle size of runway long-wave unevenness resonance fracture as described in claim 1, characterized in that, The specific process for determining the edge endpoints is as follows: The number of edge points is counted by counting the eight neighboring pixels of each edge point. Edge points with a count of 1 are recorded as edge termination points.

4. The method for measuring the particle size of runway long-wave unevenness resonance fracture as described in claim 1, characterized in that, The specific steps for obtaining the continuation line of each edge endpoint and filtering out suspected discontinuous edges are as follows: Starting from the edge endpoint, search for a continuation growth point along the connecting line; if the continuation growth point is not an edge point, continue searching for the next point along the connecting line; if the continuation growth point is an edge point, growth stops; the line connecting all the continuation growth points is recorded as the continuation line of the edge endpoint. The number of all continuous growth points on the continuous line is taken as the edge continuity length, and the continuous line with an edge continuity length greater than the preset length is taken as a suspected discontinuous edge.

5. The method for measuring the particle size of runway long-wave unevenness resonance fracture as described in claim 1, characterized in that, The process of determining the edge discontinuity confidence level of each suspected discontinuity point is as follows: Calculate the difference between the gradient magnitude of each suspected discontinuous point of each suspected discontinuous edge and the average gradient magnitude of all non-edge points in the fragmented grayscale image; Calculate the average angle between the gradient direction and the horizontal direction of all suspected discontinuous points in each suspected discontinuous edge to obtain the discontinuous gradient direction angle; Obtain the cosine of the angle between the standard vector corresponding to the end gradient direction angle of each suspected discontinuous edge and the standard vector corresponding to the discontinuous gradient direction angle; The difference between each suspected discontinuity point and the cosine value of the angle between them and the suspected discontinuity edge is positively fused to obtain the edge discontinuity confidence of each suspected discontinuity point on each suspected discontinuity edge.

6. The method for measuring the particle size of runway long-wave unevenness resonance fracture as described in claim 1, characterized in that, The process of determining the grayscale boundary balance value of each pixel is as follows: The gradient direction of each pixel is obtained as the tangent line at each pixel. The local window of each pixel is segmented using the tangent line, and the standard deviation of gray values ​​of all pixels in each segment is obtained. The negative correlation mapping result of the difference between all gray standard deviations obtained from the local window of each pixel is used as the gray value boundary balance value of each pixel.

7. The method for measuring the particle size of runway long-wave unevenness resonance fracture as described in claim 1, characterized in that, The specific process of gradient reconstruction of the fragmented grayscale image is as follows: In a fragmented grayscale image, if a pixel is a suspected discontinuity point, the gradient adjustment coefficient of the suspected discontinuity point is obtained based on the grayscale boundary balance value of the suspected discontinuity point and the edge discontinuity confidence. Through calculation The result is used to obtain the gradient magnitude after adjustment for suspected discontinuity points; where, This represents the gradient magnitude before pixel adjustment. This represents the preset first threshold. This represents the average gradient magnitude of all pixels within a preset local window of a pixel. If a pixel is an edge point, the normalized value of the grayscale boundary balance of the edge point is used as the gradient reduction coefficient of the edge point. ; When the gradient reduction coefficient is less than the preset second threshold At that time, through calculation The result is used to obtain the adjusted gradient magnitude of the corresponding edge point; otherwise, the gradient magnitude of the edge point remains unchanged. The gradient magnitude of the remaining pixels remains unchanged.

8. The method for measuring the particle size of runway long-wave unevenness resonance fracture as described in claim 7, characterized in that, The gradient adjustment coefficient is specifically the normalized value of the ratio of the edge discontinuity confidence of the suspected discontinuity point to the gray-level boundary balance value.

9. The method for measuring the particle size of runway long-wave unevenness resonance fracture as described in claim 1, characterized in that, The crushed particle size of the aggregate particles is determined by the length of the minimum bounding rectangle of the base material particle boundary.

Images