A method, device, computer device and storage medium for detecting rail abrasion

Abstract

This text provides a method, device, computer equipment and storage medium for detecting rail abrasions. The method includes: obtaining the eddy current signal of the rail to be detected; determining whether there is an abrasion on the rail and determining the position interval where the abrasion is located according to the eddy current signal and the corresponding adaptive judgment threshold; obtaining the top surface image of the rail at the position interval and performing image analysis, and determining the type of the abrasion according to the analysis result. By comparing the eddy current signal with the adaptive judgment threshold, this text excludes the influence of factors such as the environment where the rail is located and the differences in its own conductivity and magnetic permeability, and realizes the rapid and accurate detection of rail abrasions; furthermore, through the analysis of the top surface image of the position interval where the abrasion is located, the detection of the abrasion type is realized; it provides a reliable basis for comprehensively grasping the state of rail abrasions, is conducive to selecting effective treatment measures for different types of abrasions, and is conducive to improving the safety and comfort of track operation.

Description

Technical Field

[0001] The present invention relates to the technical field of rail flaw detection, and in particular to a method and device for detecting rail abrasions, a computer device, and a storage medium. Background Art

[0002] As a safe, fast, comfortable, and all-weather transportation mode, railways have become an important part of the transportation system. While the railway is developing rapidly, it also puts forward requirements for heavier and stronger rails. During the operation of the railway, poor contact between the wheels and rails is likely to cause rail damage, and rail abrasion is one of the main forms of damage. Rail abrasion will seriously affect the smoothness of the track, the stability of train operation, and the comfort of passengers. Long-term wheel-rail interaction will cause peeling and chipping of the rail or transverse fatigue cracks, and may even lead to rail fracture in severe cases, threatening the safety of train operation. Therefore, it is very important to detect rail abrasions efficiently and accurately.

[0003] The existing non-destructive detection of rail abrasions mainly uses non-destructive detection technologies and manual inspection for verification. The non-destructive detection technologies mainly include machine vision, ultrasonic, etc., but these non-destructive detection technologies all have their own limitations: for example, ultrasonic flaw detection technology can only detect abrasions and cannot evaluate the area and severity of the abrasions; the machine vision detection method can only detect rail abrasions with peeling and chipping that have occurred in the later stage of abrasion development, and cannot detect early rail abrasions without peeling and chipping. Therefore, for the detected suspected abrasions, it is necessary to combine manual inspection for verification, and manual inspection has the defects of slow detection speed and low detection efficiency.

[0004] In view of this, the purpose of this article is to provide a method and device for detecting rail abrasions, a computer device, and a storage medium, which can improve the efficiency of abrasion detection and detect the type of abrasions. Summary of the Invention

[0005] Aiming at the above problems of the prior art, the purpose of this article is to provide a method and device for detecting rail abrasions, a computer device, and a storage medium to solve the problems of low detection efficiency and incomplete detection of rail abrasions in the prior art.

[0006] To solve the above technical problems, the specific technical solutions of this article are as follows:

[0007] On the one hand, this article provides a method for detecting rail abrasions, including:

[0008] Obtaining the eddy current signal of the rail to be detected;

[0009] Determining whether the rail has abrasions and determining the position interval where the abrasions are located according to the eddy current signal and the corresponding adaptive judgment threshold;

[0010] Obtain the top surface image of the rail at the position interval and perform image analysis, and determine the type of the abrasion according to the analysis result.

[0011] Specifically, determining whether there is an abrasion on the rail and determining the position interval where the abrasion is located according to the eddy current signal and the corresponding adaptive judgment threshold includes:

[0012] Calculate the absolute value of the amplitude of the eddy current signal in each position interval;

[0013] Judge whether the absolute value of the amplitude is less than the adaptive judgment threshold;

[0014] When the amplitude is less than the adaptive judgment threshold, it is determined that there is no abrasion on the rail in the corresponding position interval;

[0015] When the amplitude is greater than or equal to the adaptive judgment threshold, it is determined that there is an abrasion on the rail in the corresponding position interval.

[0016] Preferably, the adaptive judgment threshold is obtained through the following steps:

[0017] Calculate the root mean square of the amplitude of the eddy current signal in the position interval:

[0018]

[0019] where RMS j is the root mean square of the amplitude of the eddy current signal in the j-th position interval, s k is the amplitude of the eddy current signal at the k-th detection point in the position interval, and N is the number of detection points for eddy current detection in the position interval;

[0020] According to the root mean square, calculate the adaptive judgment threshold, and the formula is:

[0021] th j = M × RMS j ;

[0022] where, th j is the adaptive judgment threshold of the j-th position interval, and M is the amplification factor.

[0023] Specifically, obtaining the top surface image of the rail at the position interval and performing image analysis, and determining the type of the abrasion according to the analysis result includes:

[0024] Divide the top surface image into multiple image sub-regions;

[0025] Calculate the mean and standard deviation of the gray level of each image sub-region;

[0026] Compare the mean value of the gray level of each of the image sub-regions with a preset first adaptive gray level threshold respectively, and compare the standard deviation of the gray level of each image sub-region with a preset second adaptive gray level threshold respectively;

[0027] When the mean value of the gray level of the image sub-region is less than or equal to the first adaptive gray level threshold, and the standard deviation of the gray level of the image sub-region is greater than or equal to the second adaptive gray level threshold, it is determined that the type of abrasion corresponding to the region enclosed by the image sub-region is chipping.

[0028] Further, before dividing the top surface image into a plurality of image sub-regions, the method further includes:

[0029] Calculate the mean value and the standard deviation of the gray level of the top surface image;

[0030] Amplify the mean value and the standard deviation of the gray level of the top surface image by coefficients respectively to obtain the first adaptive gray level threshold and the second adaptive gray level threshold.

[0031] Further, it is determined that the type of abrasion corresponding to the position of the top surface image other than the enclosed region is white layer structure.

[0032] Specifically, the obtaining the top surface image of the rail at the position interval and performing image analysis, and determining the type of abrasion according to the analysis result includes:

[0033] Calculate the saliency map of the top surface image;

[0034] Compare the gray level of each pixel point in the saliency map with a preset gray level threshold;

[0035] Obtain the saliency region composed of the pixel points whose gray level is greater than or equal to the gray level threshold, and determine that the type of rail abrasion corresponding to the position of the saliency region is chipping;

[0036] When the gray level of the pixel points are all less than the gray level threshold, it is determined that there is no saliency region in the saliency map, and the type of rail abrasion corresponding to the top surface image is white layer structure.

[0037] Further, the calculating the saliency map of the top surface image includes:

[0038] Perform Fourier transform on the top surface image to obtain the phase spectrum and the logarithmic amplitude spectrum of the top surface image:

[0039]

[0040]

[0041] Wherein, P(f) is the phase spectrum of the top surface image, I(x) is the grayscale of the top surface image at position x, and F is the Fourier transform. represents the phase; L(f) is the logarithmic amplitude spectrum of the top surface image, A(f) is the amplitude spectrum of the top surface image, A(f) = |F[I(x)]|, 丨·丨 is the amplitude value;

[0042] The logarithmic amplitude spectrum of the top surface image is filtered to obtain the residual of the logarithmic amplitude spectrum of the top surface image:

[0043]

[0044] Wherein, R(f) is the residual, is the mean filter, n 2 is a positive integer, * represents convolution operation;

[0045] Perform inverse Fourier transform and Gaussian filtering according to the residual and the phase spectrum to obtain the saliency map:

[0046] S(x)=ɡ(x)*|F -1 {exp[R(f)+iP(f)]}| 2 ;

[0047] Where S(x) is the saliency map, F -1 is the inverse Fourier transform calculation, ɡ(x) is the Gaussian low-pass filter in the spatial domain, and exp(·) is the exponential calculation with a constant e as the base.

[0048] Preferably, the grayscale threshold is obtained by the following steps:

[0049] Calculating the mean grayscale value of all pixels in the saliency map;

[0050] The grayscale threshold is obtained by amplifying the mean grayscale value of all pixels in the saliency map by a factor.

[0051] Preferably, the method further comprises:

[0052] When the mean grayscale value of the image sub-region is less than or equal to the first adaptive grayscale threshold, the standard deviation of the grayscale value of the image sub-region is greater than or equal to the second adaptive grayscale threshold, and the area enclosed by the image sub-region is a significant area corresponding to the top surface image, the type of the scratch is determined to be chipping;

[0053] Otherwise, the type of the abrasion is determined to be white layer tissue.

[0054] In a second aspect, the present invention also provides a rail scratch detection device, comprising:

[0055] An acquisition module, used for acquiring the eddy current signal of the rail to be detected;

[0056] A scratch and scratch location determination module, configured to determine whether there is a scratch on the rail and determine the location interval where the scratch is located according to the eddy current signal and the corresponding adaptive judgment threshold;

[0057] A scratch type determination module, configured to obtain the top surface image of the rail at the location interval and perform image analysis, and determine the type of the scratch according to the analysis result.

[0058] In a third aspect, the present invention also provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the method described in the above technical solution is implemented.

[0059] In a fourth aspect, the present invention also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the above technical solution is implemented.

[0060] Adopting the above technical solution, a rail scratch detection method, device, computer device and storage medium described in the present invention compare the obtained eddy current signal with the corresponding adaptive judgment threshold, eliminate the influence of factors such as the environment where different rails are located and the differences in their own conductivity and magnetic permeability, and realize the rapid and accurate detection of rail scratches; furthermore, through the analysis of the top surface image of the location interval where the scratch is located, the detection of the scratch type is realized; it provides a reliable basis for comprehensively grasping the scratch state of in-service rails, is conducive to analyzing the formation cause of the scratch, and is conducive to subsequent selection of effective repair and rectification measures for different types of scratches, and is conducive to improving the safety and comfort of track operation.

[0061] In order to make the above and other purposes, features and advantages of the present invention more obvious and understandable, the following specifically enumerates preferred embodiments and cooperates with the attached drawings to make a detailed description as follows. Description of the Drawings

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0063] Figure 1 Shows a flowchart of the steps of a rail scratch detection method provided by an embodiment of the present invention;

[0064] Figure 2The figure shows a flowchart of the steps of a method for determining whether there is a scratch and determining the location interval where the scratch is located in the embodiments of the present disclosure;

[0065] Figure 3 The figure shows a flowchart of the steps of a method for analyzing a top surface image and determining the type of scratch according to the analysis result provided by the embodiments of the present disclosure;

[0066] Figure 4 The figure shows a schematic structural diagram of an image sub-region;

[0067] Figure 5 The figure shows a flowchart of the steps of another method for analyzing a top surface image and determining the type of scratch according to the analysis result provided by the embodiments of the present disclosure;

[0068] Figure 6 The figure shows a schematic structural diagram of a rail scratch detection device provided by the embodiments of the present disclosure;

[0069] Figure 7 The figure shows a schematic structural diagram of a computer device provided by the embodiments of the present disclosure.

[0070] Description of the reference numerals in the drawings:

[0071] 61. Acquisition module;

[0072] 62. Scratch and scratch location determination module;

[0073] 63. Scratch type determination module;

[0074] 702. Computer device;

[0075] 704. Processor;

[0076] 706. Memory;

[0077] 708. Driving mechanism;

[0078] 710. Input / output module;

[0079] 712. Input device;

[0080] 714. Output device;

[0081] 716. Presentation device;

[0082] 718. Graphical user interface;

[0083] 720. Network interface;

[0084] 722. Communication link;

[0085] 724. Communication bus. Detailed implementation manners

[0086] The technical solutions in the embodiments of this article will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this article. Obviously, the described embodiments are only a part of the embodiments of this article, rather than all the embodiments. Based on the embodiments in this article, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of this article.

[0087] It should be noted that the terms "first", "second", etc. in the specification and claims of this article and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of this article described here can be implemented in an order other than those illustrated or described here. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, device, product, or equipment that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products, or equipment.

[0088] The existing methods for detecting rail abrasions mainly include non-destructive testing techniques and manual inspection and verification. The non-destructive testing techniques mainly include machine vision, ultrasound, etc. These non-destructive testing techniques provide a fast and efficient detection solution for rail detection, but these techniques all have their own limitations. For example, ultrasonic flaw detection technology can only detect abrasions and cannot evaluate the area and severity of the abrasions; the machine vision detection method can only detect rail abrasions with peeling and chipping that have occurred in the later stage of abrasion development, and cannot detect early-stage rail abrasions without peeling and chipping. Therefore, for suspected abrasions detected, it is necessary to combine manual inspection and verification, and manual inspection has the defects of slow detection speed and low detection efficiency. That is to say, the existing abrasion detection methods have the problems of low detection efficiency and inaccurate detection.

[0089] To solve the above problems, the embodiments of this article provide a method, device, computer device, and storage medium for detecting rail abrasions, which can overcome the problems of low detection efficiency and incomplete detection of rail abrasions in the prior art. Figure 1 It is a schematic diagram of the steps of a method for detecting rail abrasions provided by the embodiments of this article. This specification provides the method operation steps as described in the embodiments or flowcharts, but based on routine or non-creative labor, more or fewer operation steps may be included. The order of steps listed in the embodiments is only one way among the execution orders of numerous steps and does not represent the only execution order. When the actual system or device product is executed, it can be executed in the order of the method shown in the embodiments or the drawings or in parallel. Specifically, such as Figure 1As shown, the method may include:

[0090] S110: Obtain the eddy current signal of the rail to be detected;

[0091] In the embodiments of this specification, the eddy current signal can be detected by a probe arranged on a track inspection trolley. The detection direction of the probe is towards the rail. When the track inspection trolley moves along the track, the probe arranged on it performs eddy current detection on the top surface of the rail to obtain the eddy current signal of the rail at each detection point. The detection frequency of the probe can also be set, that is, the number of detection points for the probe to perform eddy current detection on the rail within a unit distance (for example, 1 meter) is adjusted. The higher the detection frequency, the more detection points there are within the unit distance. Specifically, the detection frequency can be set according to actual application needs. The higher the detection frequency, the more beneficial it is to improve the accuracy of scratch detection, but it will also cause a substantial increase in the amount of data, thus greatly increasing the workload of data calculation and processing.

[0092] In the embodiments of this specification, the principle of eddy current detection is as follows: The probe generates an alternating magnetic field, causing the rail to be in this alternating magnetic field and generating a swirling induced alternating current (i.e., eddy current signal) accordingly. The distribution and magnitude of the eddy current signal will be affected by factors such as the conductivity, permeability, existing defects, and defect size and shape of the rail itself. By analyzing information such as the distribution, magnitude, and phase in the eddy current signal, the defect characteristics of the detected rail can be obtained.

[0093] S120: Determine whether there is a scratch on the rail and determine the position interval where the scratch is located according to the eddy current signal and the corresponding adaptive judgment threshold;

[0094] That is, in the embodiments of this specification, each eddy current signal has a corresponding adaptive judgment threshold. Comparing each eddy current signal with its respective adaptive judgment threshold can eliminate the influence of factors such as the environment where different rails are located, their own conductivity and permeability on the discrimination of rail scratches, and improve the accuracy of scratch identification.

[0095] S130: Obtain the top surface image of the rail in the position interval and perform image analysis, and determine the type of the scratch according to the analysis result.

[0096] A rail abrasion detection method provided by an embodiment of this specification performs eddy current detection on a rail, compares the obtained eddy current signal with a corresponding adaptive judgment threshold, and realizes the automatic identification of whether there is abrasion on the rail; furthermore, it obtains the top surface image of the position interval where the abrasion is located, and determines the type of abrasion by performing image analysis on the top surface image, which not only realizes the rapid and accurate detection of rail abrasion, but also realizes the detection of the abrasion type; it provides a reliable basis for comprehensively grasping the abrasion state of in-service rails, is conducive to analyzing the formation cause of abrasion, and is conducive to subsequent selection of effective repair and rectification measures for different types of abrasions.

[0097] Specifically, as Figure 2 shown, in an embodiment of this specification, step S120: determining whether there is abrasion on the rail and determining the position interval where the abrasion is located according to the eddy current signal and the corresponding adaptive judgment threshold may further include:

[0098] S210: Calculate the absolute value of the amplitude of the eddy current signal in each position interval.

[0099] Exemplarily, the size of the position interval can be set to 1 meter. Of course, it can also be set to other sizes, and the rail is checked for abrasion in units of the position interval.

[0100] It should be noted that the setting of the position intervals can be end-to-end connected, that is, the 0-1 meter section of the rail to be detected is the first position interval, the 1-2 meter section is the second position interval, and so on; the setting of the position intervals can also be intersecting and overlapping, that is, the 0-1 meter section of the rail to be detected is the first position interval, the 0.5-1.5 meter section is the second position interval, the 1-2 meter section is the third position interval, and so on. Then, this setting method of intersecting and overlapping position intervals can realize a finer detection of the grain size of rail abrasions.

[0101] S220: Determine whether the absolute value of the amplitude is less than the adaptive judgment threshold.

[0102] In an embodiment of this specification, the adaptive judgment threshold is obtained through the following steps:

[0103] Calculate the root mean square of the amplitude of the eddy current signal in the position interval:

[0104]

[0105] where RMS j is the root mean square of the amplitude of the eddy current signal in the jth position interval, s k is the amplitude of the eddy current signal at the kth detection point in the jth position interval, and N is the number of detection points for eddy current detection in the jth position interval;

[0106] Based on the root mean square, the adaptive judgment threshold is calculated, and the formula is:

[0107] th j = M × RMS j ;

[0108] where th j is the adaptive energy threshold for the j-th position interval, M is the amplification factor, and × is the multiplication operation in the four arithmetic operations.

[0109] That is, the absolute value of the amplitude of the eddy current signals corresponding to all the detection points within the j-th position interval is compared with the adaptive judgment threshold for the j-th position interval.

[0110] S230: When the amplitude is less than the adaptive judgment threshold, it is determined that there is no abrasion on the rail within the corresponding position interval;

[0111] That is, when the absolute value of the amplitude of the eddy current signal at a certain detection point is less than the adaptive judgment threshold for the j-th position interval, it is determined that there is no abrasion on the rail corresponding to this detection point; when the absolute values of the amplitudes of the eddy current signals at all the detection points within the j-th position interval are less than the adaptive judgment threshold for the j-th position interval, it is determined that there is no abrasion on the rail corresponding to this position interval.

[0112] S240: When the amplitude is greater than or equal to the adaptive judgment threshold, it is determined that there is abrasion on the rail within the corresponding position interval.

[0113] That is to say, when the absolute value of the amplitude of the eddy current signal at a certain detection point is greater than or equal to the adaptive judgment threshold for the j-th position interval, it is determined that there is abrasion on the rail corresponding to this detection point; record the position of this detection point, which is the position of the abrasion; the position interval where this detection point is located is the position interval where the abrasion is located.

[0114] In the embodiments of this specification, by using the characteristic that the absolute value of the amplitude of the eddy current detection signal at the abrasion is significantly increased compared with the absolute value of the amplitude of the eddy current signal at the non-abraded part, the detection and positioning of the abrasion are realized, which is simple, convenient and highly accurate.

[0115] It should be noted that when the abrasion range is large, there may be a situation where the abrasion straddles multiple position intervals, and then the top surface images of the rail at these multiple position intervals are correspondingly obtained.

[0116] Since the steel rail usually extends over a long distance, a method for detecting steel rail abrasions provided in this article divides the steel rail into position intervals and then compares the eddy current signal of the steel rail within each position interval with the adaptive judgment threshold within that position interval to achieve the detection of whether there is abrasion in the steel rail within that position interval. This can eliminate the influence of the difference in the rail surface state at different positions of the same steel rail, which is beneficial to improving the accuracy of abrasion detection. Moreover, the method of detecting abrasions by dividing into position intervals can segment the long-distance steel rail and perform parallel detection on each segment, which is beneficial to improving the detection efficiency of steel rail abrasions.

[0117] In some preferred embodiments, before step S120: determining whether there is abrasion on the steel rail and determining the position interval where the abrasion is located according to the eddy current signal and the corresponding adaptive judgment threshold, the method may further include:

[0118] Perform filtering processing and denoising processing on the eddy current signal in sequence.

[0119] Performing filtering and denoising processing on the eddy current signal can eliminate the interference of factors such as different environments where the rail surface is located, different conductivity and permeability of different steel rails themselves on the eddy current signal, which is beneficial to improving the accuracy of detecting steel rail abrasions based on the eddy current signal.

[0120] In some preferred embodiments, the adaptive judgment threshold can be obtained according to the eddy current signal after filtering and denoising processing. Specifically, it includes the following steps:

[0121] Calculate the root mean square of the amplitude of the eddy current signal after filtering and denoising processing in sequence within the position interval;

[0122] Perform coefficient amplification and biasing on the root mean square to obtain the adaptive judgment threshold, that is

[0123] th j = M × RMS j + N;

[0124] where N is the bias coefficient.

[0125] It should be noted that the amplification coefficient M and the bias coefficient N can be obtained through training of the neural network on the abrasion data and impurity signal data in historical data. The selection of the amplification coefficient M and the bias coefficient N can eliminate the influence of interference signals such as rust and corrugation, so that when comparing the eddy current signal with the adaptive judgment threshold, the abrasion signal can be accurately detected.

[0126] Such as Figure 3As shown, in some feasible embodiments, step S130: obtaining the top surface image of the rail at the position interval and performing image analysis, and determining the type of the abrasion according to the analysis result, may further include:

[0127] S310: Divide the top surface image into a plurality of image sub-areas.

[0128] Preferably, the plurality of image sub-areas have the same size, and the plurality of image sub-areas can overlap each other. In the embodiment of the present specification, the sizes of the image sub-areas can be set to (2n 1 +1) pixels×(2n 1 +1) pixels; for example, Figure 4 A schematic diagram of the structure of an image sub-region with a size of 3 pixels × 3 pixels is shown. The gray value of each pixel in the image sub-region is as follows: Figure 4 The size of the image sub-region should be selected to balance the calculation accuracy and efficiency.

[0129] S320: Calculate the mean and standard deviation of the grayscale of each image sub-region.

[0130] That is, for Figure 4 The mean gray value of the image sub-area shown is:

[0131]

[0132] Among them, mean is the mean; z i,j is the gray value of the pixel in the i-th row and j-th column in the image sub-area, n 1 Is a positive integer.

[0133] For Figure 4 The standard deviation of the grayscale of the image sub-area shown is:

[0134]

[0135] Here, std represents the standard deviation.

[0136] S330: Compare the mean grayscale value of each image sub-region with a preset first adaptive grayscale threshold value, and compare the standard deviation of the grayscale value of each image sub-region with a preset second adaptive grayscale threshold value.

[0137] In the embodiment of this specification, the first adaptive grayscale threshold is:

[0138] th 1 =R 1 ×mean_all;

[0139] Among them, 1 is the first adaptive grayscale threshold, R1 is the amplification factor, R 1 is a constant; mean_all is the mean value of the gray level of the top surface image, P is the number of pixels included in each column of the top surface image, Q is the number of pixels included in each row of the top surface image, that is, the size of the top surface image is P pixels × Q pixels. That is to say, in the embodiments of this specification, the first adaptive gray level threshold is obtained by amplifying the mean value of the gray level of the top surface image by a coefficient.

[0140] Furthermore, the second adaptive gray level threshold is:

[0141] th 2 = R 2 × std_all;

[0142] where, th 2 is the second adaptive gray level threshold, R 2 is the amplification factor, R 2 is a constant; std_all is the standard deviation of the gray level of the top surface image, that is, the second adaptive gray level threshold can be obtained by amplifying the standard deviation of the gray level of the top surface image by a coefficient.

[0143] It should be noted that the values of the amplification factor R 1 and the amplification factor R 2 can be the same or different, that is, in the embodiments of this specification, the relative magnitude relationship between the amplification factor R 1 and the amplification factor R 2 is not specifically limited.

[0144] S340: When the mean value of the gray level of the image sub-region is less than or equal to the first adaptive gray level threshold, and the standard deviation of the gray level of the image sub-region is greater than or equal to the second adaptive gray level threshold, it is determined that the type of abrasion corresponding to the region enclosed by the image sub-region is chipping.

[0145] At the edge of the chipping, that is, on the demarcation line between the chipping area and the non-chipping area, the gray level of its image has an obvious difference from the gray level of the normal rail surface image; while inside the chipping area, the inner layer is exposed due to rail surface peeling, and its gray level is also different from the gray level of the normal rail surface image, but the difference is smaller and it is not easy to be accurately identified; therefore, in the embodiments of this specification, the detection of the type of abrasion is realized by first identifying the edge of the chipping and then identifying the chipping.

[0146] The mean value reflects the average level of the gray values of all pixel points in the image sub-region. When the difference between the mean value and the first adaptive gray threshold is large, it can indicate that the image sub-region is a suspected missing block area worthy of attention. However, when the difference between the mean value and the first adaptive gray threshold is small, it is also possible that the image sub-region is a suspected missing block area. Therefore, using only the mean value to measure has the problem of inaccurate detection. The standard deviation reflects the difference between the gray value of each pixel point in the image sub-region and the gray mean value. The larger the standard deviation, the greater the gray difference in the image sub-region, and the greater the possibility of missing block it shows. Therefore, in the embodiments of this specification, comparing the mean value and the standard deviation of the gray value of the image sub-region with the first adaptive gray threshold and the second adaptive gray threshold respectively can improve the accuracy of identifying the scratch type of the missing block.

[0147] For the case where the missing block edge is not recognized, that is, when the mean value and the standard deviation of the gray value of all image sub-regions in the top surface image do not meet the judgment conditions in step S340 (and the eddy current signal of the top surface image is detected to have scratches), it is determined that the scratch type of the rail corresponding to the top surface image is white layer structure.

[0148] S350: Determine that the scratch type corresponding to the position other than the area enclosed by the image sub-region in the top surface image is white layer structure.

[0149] The white layer structure is formed due to the high temperature generated by the friction between the train wheel and the top surface of the rail. At the same time, the high contact stress reduces the phase transformation temperature of the material, resulting in the phase transformation of the metal structure on the top surface of the rail head, and then it is transformed from the pearlite structure. The white layer structure has high hardness, high brittleness, and poor toughness. Under continuous external force loading, the white layer structure breaks and fractures, thus forming a missing block. It can be seen that the white layer structure is an early manifestation of scratches; while the missing block is the main manifestation form of late scratches. When a missing block occurs on the rail surface, it can be considered that the scratch degree is relatively serious at this time.

[0150] In fact, for the top surface image of the rail with a missing block, it is very likely that the area other than the missing block is a potential missing block area (there is also a small possibility that it is a normal rail surface), that is, the white layer structure has been formed but has not developed to a missing block. However, with the further contact between the wheel and the rail, the area of the missing block area will become larger and larger. Therefore, in the embodiments of this specification, the scratch type of the area other than the missing block area in the top surface image is determined as white layer structure, so as to enable the inspectors to strengthen the supervision and renovation of it.

[0151] It should be noted that the embodiments of this specification provide formulas for calculating the mean and standard deviation of the gray levels of an image sub-region where the number of pixels in each column and each row is odd. In actual application scenarios, the number of pixels in each column or each row of the image sub-region can also be even. For the calculation of the mean and standard deviation of the gray levels of the image sub-region in these cases, those skilled in the art can obtain them by referring to the above calculation formulas, and will not be elaborated here.

[0152] A rail abrasion detection method provided by the embodiments of this specification utilizes the characteristic that there is an obvious gray level difference between the top surface image of a rail with a chip and the top surface image of a normal rail to detect rail abrasions manifested as chips; and the abrasions detected by the eddy current detection method with no obvious gray level difference are abrasions of the white layer tissue type; thus, the detection of the abrasion type is realized, which is beneficial to the discovery of early abrasions and convenient for the effective prevention, treatment and maintenance of early abrasions. Moreover, through the above image analysis method, an image sub-region with an obvious gray level difference is detected, and the accurate positioning of the chip-type abrasion can be achieved by positioning the position of the image sub-region on the top surface image.

[0153] As Figure 5 shown, in some other feasible embodiments, step S130: Obtain the top surface image of the rail in the position interval and perform image analysis, and determine the type of the abrasion according to the analysis result, which can further include:

[0154] S510: Calculate the saliency map of the top surface image;

[0155] Denote the saliency map as S(x), where x is a pixel point in the top surface image. The saliency map corresponds to the top surface image, but the gray level difference between the pixel points in it is larger than that between the pixel points in the top surface image. Thus, it is beneficial to the screening of chip-type abrasions.

[0156] S520: Compare the gray levels of the pixel points in the saliency map with a preset gray level threshold;

[0157] S530: Obtain the saliency region composed of the pixel points whose gray levels are greater than or equal to the gray level threshold, and determine that the type of the rail abrasion at the corresponding position of the saliency region is a chip;

[0158]

[0159] where T is the gray level threshold, and the region where Z(x)=1 is the saliency region. That is, on the basis of the saliency map, the gray levels of the pixel points are binarized, so that the difference between the saliency region and other regions is more obvious, which is convenient for the calibration of different types of abrasions and the abrasion positions.

[0160] Compare the grayscale of each pixel point in the saliency map with the grayscale threshold, record the pixel points that satisfy the condition that the grayscale is greater than or equal to the grayscale threshold and form a set, and the region represented by this set is the saliency region.

[0161] S540: When the grayscale of the pixel points is less than the grayscale threshold, it is determined that there is no saliency region in the saliency map, and the type of rail abrasion corresponding to the top surface image is white layer structure.

[0162] Since abrasions of the early white layer structure type are difficult to be recognized by the human eye, and the method for analyzing the rail top surface image based on the saliency map provided in the embodiments of this specification to detect the abrasion type can detect early white layer structure type abrasions and late chipping type abrasions, greatly reducing the working pressure and workload of manual inspection, and greatly improving the accuracy and detection efficiency of abrasion detection.

[0163] In some specific embodiments, the grayscale threshold T can be obtained by calculating the mean value of the grayscale of all pixel points in the saliency map and performing coefficient amplification.

[0164] Specifically, step S510: Calculating the saliency map of the top surface image includes the following steps:

[0165] Perform Fourier transform on the top surface image to obtain the phase spectrum and logarithmic amplitude spectrum of the top surface image:

[0166]

[0167]

[0168] Among them, P(f) is the phase spectrum of the top surface image, I(x) is the grayscale of the top surface image at x, x is the pixel point in the top surface image, F is the Fourier transform, represents obtaining the phase; L(f) is the logarithmic amplitude spectrum of the top surface image, A(f) is the amplitude spectrum of the top surface image, A(f) = |F[I(x)]|, and |·| represents taking the amplitude value.

[0169] Perform filtering processing on the logarithmic amplitude spectrum of the top surface image to obtain the residual of the logarithmic amplitude spectrum of the top surface image:

[0170]

[0171] Among them, R(f) is the residual, is the mean filter, and * represents the convolution operation; preferably, is an n 2 ×n 2 mean filter (n2 is a positive integer, such as n 2 takes a value of 3), and the expression is as follows:

[0172]

[0173] Smoothing L(f) to obtain the smoothed logarithmic amplitude spectrum. The R(f) obtained by subtracting L(f) from the smoothed logarithmic amplitude spectrum (referred to as the residual) is the region smoothed in the frequency domain, that is, the significant region.

[0174] Performing inverse Fourier transform and Gaussian filtering according to the residual and the phase spectrum to obtain the significant map:

[0175] S(x) = ɡ(x) * |F -1 {exp[R(f) + iP(f)]}| 2 ;

[0176] where S(x) is the significant map, F -1 is the inverse Fourier transform calculation, ɡ(x) is the Gaussian filter in the spatial domain, i represents the imaginary part, and exp(·) is the exponential calculation with the constant e as the base.

[0177] That is, the information in the frequency domain is transformed into the image information in the spatial domain through the inverse Fourier transform to obtain the significant map of the top surface image, and better display effects are achieved through Gaussian filtering.

[0178] According to the foregoing, it can be known that there are obvious gray-scale differences between the top surface images of the rails with block shedding and the top surface images of normal rails, and the gray-scale information of the top surface images also exists in the frequency spectra obtained by their Fourier transforms. The logarithmic amplitude spectra of most images (images without block shedding) have approximately the same shape. When the part different from the same shape in the logarithmic amplitude spectrum of the image (i.e., the residual part of the logarithmic amplitude spectrum) contains new information, it is the part that this specification embodiment focuses on (i.e., the significant region characterized as block-shedding type abrasion). In this specification embodiment, this theory is used to perform significant analysis on the top surface images of the rails detected with abrasion by eddy current testing, detect block shedding and white layer tissues, and realize the identification of abrasion types.

[0179] In some other feasible embodiments, the analysis results of the image sub-region gray-scale analysis method and the significant region analysis method can be combined to realize the detection of abrasion types. Specifically, it can be:

[0180] When the mean value of the gray scale of the image sub-region is greater than or equal to the first adaptive gray-scale threshold, the standard deviation of the gray scale of the image sub-region is greater than or equal to the second adaptive gray-scale threshold, and the image sub-region is the significant region corresponding to the top surface image, it is determined that the type of the abrasion is block shedding;

[0181] Otherwise, the type of the abrasion is determined to be white layer tissue.

[0182] By combining the above two image analysis methods, the accuracy of detecting scratch types can be improved.

[0183] Preferably, in the embodiment of this specification, after determining the type of the abrasion according to the analysis result in step S130, the method may further include:

[0184] Depending on the type of abrasion in question, choose an appropriate remediation measure.

[0185] For example, for abrasions determined to be block falling, the rails can be polished for maintenance and repair; for abrasions determined to be white layer tissue, long-term follow-up detection measures can be taken.

[0186] In summary, the rail scratch detection method provided in the embodiments of the present specification can eliminate the influence of factors such as the environment in which the rail is located, the difference in the electrical conductivity and magnetic permeability of the rail itself, etc. on the scratch identification, realize automatic detection of rail scratches, and improve the efficiency and accuracy of scratch detection; and accurately detect the scratches and the locations of the scratches; and use the characteristic that there is an obvious grayscale difference between the top surface image of the rail with broken blocks and the top surface images of the rail at other positions to detect different types of scratches, which can be used to analyze the severity of the scratches, and can select subsequent rail maintenance and remediation measures in a targeted manner, thereby improving the safety and comfort of track operation.

[0187] like Figure 6 As shown, the embodiment of this specification also provides a rail scratch detection device, including:

[0188] An acquisition module 61 is used to acquire an eddy current signal of a rail to be detected;

[0189] A scratch and scratch position determination module 62, which determines whether the rail has scratches and determines the position interval of the scratches according to the eddy current signal and the corresponding adaptive judgment threshold;

[0190] The scratch type determination module 63 obtains the top surface image of the rail at the position interval and performs image analysis, and determines the type of the scratch according to the analysis result.

[0191] The beneficial effects obtained by the device provided in the embodiments of this specification are consistent with the beneficial effects obtained by the above method and will not be repeated here.

[0192] like Figure 7As shown, a computer device provided by an embodiment of this article. The computer device 702 may include one or more processors 704, such as one or more central processing units (CPUs), and each processing unit may implement one or more hardware threads. The computer device 702 may also include any memory 706, which is used to store any kind of information such as code, settings, data, etc. Non-limiting examples include any one or a combination of the following: any type of RAM, any type of ROM, flash memory devices, hard disks, optical discs, etc. More generally, any memory may use any technology to store information. Further, any memory may provide volatile or non-volatile retention of information. Further, any memory may represent a fixed or removable component of the computer device 702. In one case, when the processor 704 executes the associated instructions stored in any memory or combination of memories, the computer device 702 may perform any operation of the associated instructions. The computer device 702 also includes one or more drive mechanisms 708 for interacting with any memory, such as a hard disk drive mechanism, an optical disc drive mechanism, etc.

[0193] The computer device 702 may also include an input / output module 710 (I / O), which is used to receive various inputs (via the input device 712) and to provide various outputs (via the output device 714). A specific output mechanism may include a presentation device 716 and an associated graphical user interface (GUI) 718. In other embodiments, the input / output module 710 (I / O), the input device 712, and the output device 714 may not be included, and it may only be a computer device in a network. The computer device 702 may also include one or more network interfaces 720, which are used to exchange data with other devices via one or more communication links 722. One or more communication buses 724 couple the components described above together.

[0194] The communication link 722 may be implemented in any way, for example, through a local area network, a wide area network (e.g., the Internet), a point-to-point connection, etc., or any combination thereof. The communication link 722 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc. governed by any protocol or combination of protocols.

[0195] Corresponding to Figures 1 to 3 and Figure 5 For the method in, embodiments of this article also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, it executes the steps of the above method.

[0196] The embodiments of the present disclosure also provide a set of computer-readable instructions. When a processor executes the instructions, the program therein causes the processor to execute the method as shown in Figures 1 to 3 and Figure 5 .

[0197] It should be understood that in various embodiments of the present disclosure, the sequence numbers of the above processes do not imply the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present disclosure.

[0198] It should also be understood that in the embodiments of the present disclosure, the term "and / or" is merely a description of the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in the present disclosure generally represents an "or" relationship between the associated objects before and after.

[0199] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present disclosure.

[0200] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0201] In the several embodiments provided in the present disclosure, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed couplings or direct couplings or communication connections to each other can be indirect couplings or communication connections through some interfaces, devices, or units, and can also be electrical, mechanical, or other forms of connections.

[0202] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, that is, it may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiments of this article.

[0203] In addition, each functional unit in the various embodiments of this article may be integrated into a processing unit, may exist separately as individual physical units, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0204] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this article, in essence, or the part that contributes to the prior art, or all or part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this article. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0205] Specific embodiments are used in this article to elaborate on the principles and implementation manners of this article. The descriptions of the above embodiments are only used to help understand the method and its core idea of this article; at the same time, for those of ordinary skill in the art, according to the idea of this article, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to this article.

Claims

1. A method for detecting rail abrasion, characterized in that, it includes: Obtaining the eddy current signal of the rail to be detected; According to the eddy current signal and the corresponding adaptive judgment threshold, determining whether there is abrasion on the rail and determining the position interval where the abrasion is located; wherein, the adaptive judgment threshold is obtained through the following steps: Calculating the root mean square of the amplitude of the eddy current signal within the position interval: Among them, RMS j is the root mean square of the eddy current signal amplitude within the j-th position interval, s k is the eddy current signal amplitude at the k-th detection point within the position interval, and N is the number of detection points for eddy current detection within the position interval; According to the root mean square, calculating the adaptive judgment threshold, and the formula is: th j = M × RMS j ; where th j is the adaptive judgment threshold for the j-th position interval, and M is the amplification factor; Obtaining the top surface image of the rail at the position interval and performing image analysis, and determining the type of the abrasion according to the analysis result.

2. The method according to claim 1, characterized in that, The determining whether there is abrasion on the rail and determining the position interval where the abrasion is located according to the eddy current signal and the corresponding adaptive judgment threshold includes: Calculating the absolute value of the amplitude of the eddy current signal within each position interval; Judging whether the absolute value of the amplitude is less than the adaptive judgment threshold; When the amplitude is less than the adaptive judgment threshold, it is determined that there is no abrasion on the rail within the corresponding position interval; When the amplitude is greater than or equal to the adaptive judgment threshold, it is determined that there is abrasion on the rail within the corresponding position interval.

3. The method according to claim 2, characterized in that, The obtaining the top surface image of the rail at the position interval and performing image analysis, and determining the type of the abrasion according to the analysis result further includes: Dividing the top surface image into multiple image sub-regions; Calculating the mean value and standard deviation of the gray level of each image sub-region; Comparing the mean value of the gray level of each image sub-region with a preset first adaptive gray level threshold respectively, and comparing the standard deviation of the gray level of each image sub-region with a preset second adaptive gray level threshold respectively; When the mean value of the gray level of the image sub-region is less than or equal to the first adaptive gray level threshold, and the standard deviation of the gray level of the image sub-region is greater than or equal to the second adaptive gray level threshold, it is determined that the type of the abrasion corresponding to the region enclosed by the image sub-region is spalling.

4. The method according to claim 3, characterized in that, Before dividing the top surface image into multiple image sub-regions, the method further includes: Calculating the mean value and standard deviation of the gray level of the top surface image; Performing coefficient amplification on the mean value and standard deviation of the gray level of the top surface image respectively to obtain the first adaptive gray level threshold and the second adaptive gray level threshold.

5. The method according to claim 3, characterized in that, The method further includes: Determining that the type of the abrasion corresponding to the position other than the region enclosed by the top surface image is white layer structure.

6. The method according to claim 2, characterized in that, The obtaining the top surface image of the rail at the position interval and performing image analysis, and determining the type of the abrasion according to the analysis result further includes: Calculating the saliency map of the top surface image; Comparing the gray level of each pixel point in the saliency map with a preset gray level threshold; Obtain a saliency region composed of the pixel points whose gray scale is greater than or equal to the gray scale threshold, and determine that the type of the rail abrasion at the corresponding position of the saliency region is a chip; When the gray scale of all the pixel points is less than the gray scale threshold, it is determined that there is no saliency region in the saliency map, and the type of the rail abrasion corresponding to the top surface image is white layer structure.

7. The method according to claim 6, characterized in that, the calculating the saliency map of the top surface image further includes: Performing Fourier transform on the top surface image to obtain the phase spectrum and the logarithmic amplitude spectrum of the top surface image: Wherein, P(f) is the phase spectrum of the top surface image, I(x) is the grayscale of the top surface image at position x, and F is the Fourier transform. represents the phase; L(f) is the logarithmic amplitude spectrum of the top surface image, A(f) is the amplitude spectrum of the top surface image, A(f) = |F[I(x)]|, |F[I(x)] is the amplitude of F[I(x)]; Performing filtering processing on the logarithmic amplitude spectrum of the top surface image to obtain the residual of the logarithmic amplitude spectrum of the top surface image: where R(f) is the residual error, is a mean filter, and n 2 is a positive integer, and * represents the convolution operation; Performing inverse Fourier transform and Gaussian filtering according to the residual and the phase spectrum to obtain the saliency map: S(x) = g(x) * F -1 {exp[R(f) + iP(f)]} 2 ; where S(x) is the saliency map, F -1 is the inverse Fourier transform calculation, ɡ(x) is the Gaussian filter, and exp(·) is the exponential calculation with the constant e as the base.

8. The method according to claim 6, characterized in that, the gray scale threshold is obtained through the following steps: Calculating the mean value of the gray scale of all pixel points in the saliency map; Performing coefficient amplification on the mean value of the gray scale of all pixel points in the saliency map to obtain the gray scale threshold.

9. The method according to claim 4 or claim 6, characterized in that, the method further includes: When the mean value of the gray scale of the image sub-region is less than or equal to the first adaptive gray scale threshold, the standard deviation of the gray scale of the image sub-region is greater than or equal to the second adaptive gray scale threshold, and the region enclosed by the image sub-region is the saliency region corresponding to the top surface image, it is determined that the type of the abrasion is a chip; Otherwise, it is determined that the type of the abrasion is white layer structure.

10. A rail abrasion detection device, characterized in that, including: An acquisition module, configured to acquire the eddy current signal of the rail to be detected; An abrasion and abrasion position determination module, configured to determine whether there is an abrasion on the rail and determine the position interval where the abrasion is located according to the eddy current signal and the corresponding adaptive judgment threshold; wherein, the adaptive judgment threshold is obtained through the following steps: Calculating the root mean square of the amplitude of the eddy current signal in the position interval: Among them, RMS j is the root mean square of the eddy current signal amplitude within the j-th position interval, s k is the eddy current signal amplitude of the k-th detection point within the position interval, and N is the number of detection points for eddy current detection within the position interval; Calculating the adaptive judgment threshold according to the root mean square, and the formula is: th j = M × RMS j ; where th j is the adaptive judgment threshold of the j-th position interval, and M is the amplification factor; An abrasion type determination module, configured to acquire the top surface image of the rail at the position interval and perform image analysis, and determine the type of the abrasion according to the analysis result.

11. A computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, when the processor executes the computer program, the method described in any one of claims 1 to 9 is implemented.

12. A computer-readable storage medium, characterized in that, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any one of claims 1 to 9 is implemented.

Images