Steel wire rope surface damage detection method and detection device

Abstract

The invention relates to a steel wire rope surface damage detection method and detection device.The steel wire rope surface damage detection method comprises the following steps that a binocular vision detection device is arranged, the binocular vision detection device comprises two cameras, and the positions of the cameras and a light source are adjusted so that the surface of a steel wire rope can be clearly shot; acquiring an image of the surface of the steel wire rope through a camera, and transmitting the image to a computer; preprocessing the image to at least improve the quality and contrast of the image; feature extraction is carried out on the preprocessed image to extract the features of the surface damage of the steel wire rope, and feature extraction comprises edge detection, contour extraction and texture analysis; analyzing the extracted features by adopting a damage identification algorithm, and judging whether the surface of the steel wire rope is damaged or not; outputting a detection result of the surface damage of the steel wire rope, wherein the detection result comprises damage types, damage positions and damage degrees. The detection method can automatically identify the surface damage of the steel wire rope, reduces manual intervention, and improves the detection efficiency and accuracy.

Description

Technical Field

[0001] The present application relates to the technical field of nondestructive testing, and in particular to a method and device for detecting surface damage of a steel wire rope. Background Art

[0002] Wire rope is a key component in many mechanical equipment, such as cranes, elevators, ropeways, etc. During use, the wire rope may be damaged by wear, breakage, corrosion, etc., which may cause serious accidents. Therefore, it is very important to regularly inspect and maintain the wire rope. At present, the commonly used wire rope inspection methods include visual inspection, magnetic particle inspection, ultrasonic inspection, etc. However, these methods have some limitations. Visual inspection requires manual operation, is inefficient, and is easily affected by the subjective factors of the inspectors; magnetic particle inspection and ultrasonic inspection require pre-treatment of the wire rope, which is complicated to operate and has high cost. Summary of the invention

[0003] A first aspect of an embodiment of the present application provides a method for detecting surface damage of a wire rope, which can automatically identify damage on the surface of the wire rope, reduce manual intervention, and improve detection efficiency and accuracy.

[0004] The first aspect of the present invention provides a method for detecting surface damage of a steel wire rope, comprising the following steps:

[0005] 1) Arrange a binocular vision detection device, which includes at least two cameras, and adjust the positions of the cameras and the light source so that the surface of the wire rope can be clearly photographed;

[0006] 2) collecting an image of the surface of the steel wire rope by the camera and transmitting the image to a computer;

[0007] 3) preprocessing the image to at least improve the quality and contrast of the image;

[0008] 4) performing feature extraction on the preprocessed image to extract features of the surface damage of the wire rope, wherein the feature extraction includes edge detection, contour extraction and texture analysis;

[0009] 5) Using a damage identification algorithm to analyze the extracted features to determine whether there is damage on the surface of the wire rope;

[0010] 6) Outputting the detection result of the surface damage of the steel wire rope, wherein the detection result includes the damage type, damage location and damage degree.

[0011] In addition, the wire rope surface damage detection method provided in the embodiment of the present application may also have the following additional technical features:

[0012] In an optional solution, in step 3), the preprocessing of the image includes homomorphic filtering to remove noise in the image, which specifically includes the following steps:

[0013] Assume that a digital image is g(x,y), its incident component is u(x,y), and its reflected component is v(x,y), then:

[0014] g(x,y)=u(x,y)·v(x,y);

[0015] Take the logarithm of both sides:

[0016] lng(x,y)=lnu(x,y)+lnv(x,y);

[0017] Take the Fourier transform of both sides:

[0018] G(u,v)=U(u,v)+V(u,v);

[0019] Suppose that the filter function H(u,v) is used to process G(u,v):

[0020] H(u,v)G(u,v)=H(u,v)U(u,v)+H(u,v)V(u,v);

[0021] Then switch back to the airspace:

[0022] h g (x,y)=h u (x,y)+h v (x,y);

[0023] Taking the exponents on both sides gives the homomorphic filtering result:

[0024]

[0025] In an optional solution, in step 3), the preprocessing of the image includes binarization to adaptively enhance the image, which specifically includes the following steps:

[0026] The image is considered to be composed of an incident image and a reflected image. The incident light shines on the reflective object, and the reflected light enters the human eye through reflection by the reflective object. The final image can be expressed by the following formula:

[0027] I(x,y)=L(x,y)*R(x,y);

[0028] Where: I(x,y) represents the image signal observed or received by the camera, L(x,y) represents the illumination component of the ambient light, and R(x,y) represents the reflection component of the target object carrying the image detail information;

[0029] Taking the logarithm of both sides of the equation gives:

[0030] Log(R(x,y))=Log(I(x,y))-Log(L(x,y));

[0031] L(x,y) is approximated by the convolution of I(x,y) and a Gaussian kernel:

[0032] L(x,y)=I(x,y)*G(x,y);

[0033] In the above formula, * represents convolution, G(x,y) represents Gaussian kernel;

[0034] Finally, Log(R(x,y)) is quantized to pixel values ​​in the range of 0-255:

[0035] R(x,y)=(Value-Min) / (Max-Min)*255.

[0036] In an optional solution, in the step 2), when collecting the image of the wire rope surface, it is necessary to clarify the relationship between the camera parameters and the wire rope parameters, including the relationship between the parameters when collecting the full rope information, the relationship between the actual field of view and the camera imaging surface, and the relationship between the camera parameters and the wire rope speed during high-speed operation;

[0037] Among them, the relationship between the running speed of the wire rope, the running time, the camera frame rate, the actual field of view height, and the overlap of adjacent images is: Fh-(F-1)s=vt; where F is the camera frame rate, s is the overlap between two adjacent images, v is the running speed of the wire rope, t is the running time of the wire rope, and h is the actual field of view height of the wire rope;

[0038] The basic parameters of the camera and lens: S = h × l, so: F (h × l × D) - f (F-1) s = vtf; where W is the actual optical magnification, f is the focal length, D is the object distance, S is the imaging height, h resolution is the height of the image resolution, l pixel is the size of the camera pixel, and h is the actual field of view height of the wire rope.

[0039] In an optional solution, in the step 2), when performing image acquisition on the surface of the wire rope, the acquisition strategy used is to establish a camera class in the algorithm, and coordinate resources on the basis of multi-camera instantiation through a multi-threaded parallel acquisition structure;

[0040] The data storage strategy used is as follows: after the system and program are started, the algorithm starts three threads to process the data and position data of the two cameras respectively, and the computer saves while collecting; when storing, a cache area is first opened in the memory, which consists of at least 4 one-dimensional arrays, each of which can store 300 frames of images, occupying 9G of memory and 32G of system memory; when a cache area is filled, the save flag flag+1 is saved, the saving thread starts saving from the first array, and the acquisition thread continues to collect and save to the first array, and so on;

[0041] When the acquisition thread stops, the saving thread will continue to save until the images in the cache area are emptied, and then use avi video encoding technology to encode the images to form a video, and finally save the video and location data to the hard disk and txt file respectively.

[0042] In an optional solution, in step 4), an edge detection algorithm is used to extract edge features of the surface of the steel wire rope, and the edge detection algorithm determines the position of the edge by calculating the gradient of the image, thereby finding the contour of the steel wire rope and possible damaged areas;

[0043] The contour extraction is performed by using a contour extraction function in an image processing library, wherein the contour extraction function includes a findContours function in OpenCV;

[0044] When performing the texture analysis, the content of the image is described by analyzing the texture features of the image, and the texture features include gray-level co-occurrence matrix and local binary pattern.

[0045] In an optional solution, the support vector machine algorithm is used in step 5) to classify the extracted features to determine whether there is damage on the surface of the wire rope, and the analysis of the extracted features includes:

[0046] The video calculation method is used to obtain the damaged frame time: first select the video, including the date, wire rope number and camera position, and then perform OpenCV decoding to obtain the video frame rate, encoding method, image height and width information; then, the system detects each frame of the image, and if the wire rope is damaged, the formula is used to calculate the damage frame time. Calculate the time of the damaged frame and find the corresponding position in the txt file table. In the formula, k is the frame number of the damaged frame, f2 is the frame rate of the read video, and t2 is the time of the damaged frame.

[0047] Removal of repeated damage: After the video calculation results are output, when the first damage is detected, x0 and h0 are recorded, i represents the number of damages, and i is 1 at this time; when the second damage is detected, x1 and h1 are recorded, and then the algorithm enters the formula: |(x1-x0)*c+(h1-h0)|>T, and makes a judgment. If the conditions are met, the damage is saved, and i, which counts the number of damages, is added by 1; otherwise, h1 is assigned to h0, x1 is assigned to x0, and then the next detection is performed until all images are detected; in the above formula, c represents the ratio between the pixel size in the image and the actual size, and T represents the error size.

[0048] In an optional solution, in step 1), adjusting the camera includes adjusting camera parameters, specifically including:

[0049] Adjust the camera's focus, aperture, shutter speed, and adjust camera parameters according to specific environment and needs;

[0050] Initialize the parameters, including setting the camera frame rate to 100 frames to ensure that sufficient image details can be captured, setting the resolution to 1280×1024 to obtain a clear image, and setting the saved video format and file address for subsequent storage and processing of image data;

[0051] Start the camera and open three threads, including two camera threads and one position data acquisition thread. The two camera threads are used to collect image data from two cameras at the same time to realize the binocular vision function. The position data acquisition thread is used to obtain the camera's position information for subsequent image processing and analysis.

[0052] A second aspect of an embodiment of the present application provides a wire rope surface damage detection device, the wire rope surface damage detection device comprising:

[0053] A binocular vision detection device, the binocular vision detection device comprising at least two cameras, the two cameras being capable of simultaneously capturing images of the surface of the wire rope;

[0054] A light source, the light source is used to illuminate the surface of the wire rope;

[0055] An image acquisition card, wherein the image acquisition card is used to acquire images taken by the camera;

[0056] An image processing module and a computer, wherein the image processing module is used for preprocessing, feature extraction and damage identification of images, and the computer is used for processing the images collected by the image acquisition card and outputting the detection results of the surface damage of the wire rope.

[0057] In an optional solution, the camera is an industrial camera, the light source is an LED light source, and the image acquisition card is a high-speed image acquisition card.

[0058] The beneficial effects of the embodiments of the present application are:

[0059] The wire rope surface damage detection method in the embodiment of the present application adopts binocular vision technology to detect the surface damage of the wire rope, which can automatically identify the damage on the surface of the wire rope, reduce human intervention, improve the detection efficiency and accuracy, and has the advantages of non-contact, high efficiency and high precision. In addition, the wire rope surface damage detection device using this method has the advantages of simple structure, easy operation and low cost.

[0060] It should be understood that the foregoing general description and the following detailed description are exemplary only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A schematic diagram of the detection process of the wire rope surface damage detection method provided in this application;

[0062] Figure 2 Schematic diagram of the wire rope collection method provided for this application;

[0063] Figure 3 A schematic diagram of the homomorphic filtering algorithm flow provided for this application;

[0064] Figure 4 This is a schematic diagram of the structure of the wire rope surface damage detection device provided in this application.

[0065] Reference numerals: camera 1 , light source 2 , image acquisition card 3 , computer 4 , image processing module 5 , 220V power supply 6 .

[0066] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application. DETAILED DESCRIPTION

[0067] In order to better understand the technical solution of the present application, the embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0068] It should be clear that the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0069] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms "a", "said" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0070] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0071] It should be noted that the directional words such as "upper", "lower", "left", and "right" described in the embodiments of the present application are described at the angles shown in the accompanying drawings and should not be understood as limiting the embodiments of the present application. In addition, in the context, it is also necessary to understand that when it is mentioned that an element is connected to another element "upper" or "lower", it can not only be directly connected to another element "upper" or "lower", but also indirectly connected to another element "upper" or "lower" through an intermediate element.

[0072] like Figure 1-4 As shown, the first aspect of the embodiment of the present application provides a method for detecting surface damage of a steel wire rope, which uses binocular vision technology to detect surface damage of the steel wire rope, can automatically identify damage on the surface of the steel wire rope, reduces manual intervention, improves detection efficiency and accuracy, and has the advantages of non-contact, high efficiency, high precision, etc. Specifically, the method for detecting surface damage of a steel wire rope includes the following steps:

[0073] 1) Arrange a binocular vision detection device, which includes at least two cameras 1, and adjust the positions of the cameras 1 and the light source 2 so that the surface of the wire rope can be clearly photographed;

[0074] 1-1 Install the camera: Carefully install the camera 1 in the appropriate position, and adjust the angle and height of the camera 1 to ensure that it can clearly capture the surface of the wire rope. During the installation process, it is necessary to avoid external interference to the camera, such as vibration, light, etc., to ensure the quality of the captured image. In addition, it is necessary to ensure that the distance between the camera and the wire rope is appropriate so that the detailed information of the wire rope surface can be captured.

[0075] 1-2 Adjust camera parameters: According to the actual situation, adjust the focal length, aperture, shutter speed and other parameters of camera 1 to obtain the best image quality. First, initialize the parameters, which includes setting the camera frame rate to 100 frames to ensure that sufficient image details can be captured; at the same time, set the resolution to 1280×1024 to obtain a clear image. In addition, it is necessary to set the saved video format and file address so that the image data can be stored and processed later.

[0076] Secondly, start camera 1 and open three threads, including two camera threads and one position data acquisition thread. The two camera threads are used to simultaneously collect image data from two cameras to realize the binocular vision function, and the position data acquisition thread is used to obtain the camera's position information for subsequent image processing and analysis.

[0077] In actual applications, it is necessary to further adjust the camera parameters according to the specific environment and needs. For example, in dim light, the aperture can be increased or the shutter speed can be extended to increase the amount of light entering; when shooting fast-moving objects, the shutter speed can be increased to avoid image blur, and the focal length can be adjusted to achieve clear imaging of objects at different distances.

[0078] 1-3 Continuous shooting: After starting the camera, it will continuously shoot the wire rope to obtain a series of images of the wire rope surface. These images will form an image sequence, which contains the surface conditions of the wire rope at different time points. By analyzing this image sequence, you can have a more comprehensive understanding of the surface conditions of the wire rope and detect whether there are damages or other abnormalities. This continuous shooting method can improve the accuracy and reliability of detection and help to discover potential problems with the wire rope in a timely manner. It should be noted that when shooting continuously, the camera settings and shooting conditions should be kept consistent to obtain comparable and accurate image data. In addition, the need to store and process large amounts of image data should be considered to ensure efficient operation of the system.

[0079] 2) Capture the image of the wire rope surface through the camera and transmit the image to the computer;

[0080] 2-1 The high-speed acquisition strategy adopted by this application is to establish a camera class in the algorithm, and through a multi-threaded parallel acquisition structure, coordinate resources based on multi-camera instantiation. Image acquisition mainly uses the collector GetImg to capture images. The collector GetImg has two functions. The first function is to capture images in sequence, and the second function is to capture specific images.

[0081] 2-2 Data storage strategy: During the data storage process, after the system and program are started, the algorithm starts three threads to process the data of the first camera, the data of the second camera and the position data respectively, and the computer saves while collecting. When storing, a cache area is first opened in the memory. The area consists of 4 one-dimensional arrays. Each array can store 300 frames of images, occupying 9G memory and 32G system memory, which can meet the needs. When a cache area is filled, the flag flag+1 is saved, and the saving thread starts saving from the first array. The acquisition thread continues to collect and save to the first array, and so on. When the acquisition thread stops, the saving thread will continue to save until the image in the cache area is emptied, and then the image is encoded using avi video encoding technology to form a video. Finally, the video and position data are saved to the hard disk and txt file respectively.

[0082] 2-3 Wire rope acquisition method: Clear acquisition of wire rope images is the core of the visual inspection system. For wire ropes running at high speeds and with small damage targets, the relationship between the visual inspection system parameters and the wire rope parameters must be clarified. Figure 2 A schematic diagram of the wire rope acquisition method is shown, including the relationship between various parameters when acquiring full rope information, the relationship between the actual field of view and the camera imaging surface, and the relationship between camera parameters and wire rope speed during high-speed operation.

[0083] The relationship between the wire rope running speed, running time and the camera frame rate, actual field of view height, and the overlap of adjacent images in the visual inspection system parameters is: Fh-(F-1)s=vt; where F is the camera frame rate, s is the overlap between two adjacent images, v is the running speed of the wire rope, t is the running time of the wire rope, and h is the actual field of view height of the wire rope.

[0084] The basic parameters of the camera and lens: S=h×l, so: F(h×l×D)-f(F-1)s=vtf, where W is the actual optical magnification, f is the focal length, D is the object distance, S is the imaging height, h resolution is the height of the image resolution, l pixel is the size of the camera pixel, and h is the actual field of view height of the wire rope; from the formula, it can be seen that the larger the focal length, the smaller the actual field of view height of the wire rope; the larger the object distance, the larger the actual field of view height of the wire rope.

[0085] 3) Preprocessing the image to at least improve the quality and contrast of the image, the preprocessing including grayscale, filtering, binarization and other operations;

[0086] 3-1 Denoising: Use homomorphic filtering method to preprocess the image to enhance the details of the image without losing the texture details of the bright area. The flowchart is as follows Figure 3As shown. Assuming a digital image is g(x,y), its incident component is u(x,y), and its reflected component is v(x,y), then the following formula can be obtained:

[0087] g(x,y)=u(x,y)·v(x,y);

[0088] Take the logarithm of both sides:

[0089] lng(x,y)=lnu(x,y)+lnv(x,y);

[0090] Take the Fourier transform of both sides:

[0091] G(u,v)=U(u,v)+V(u,v);

[0092] Suppose that the filter function H(u,v) is used to process G(u,v):

[0093] H(u,v)G(u,v)=H(u,v)U(u,v)+H(u,v)V(u,v);

[0094] Then switch back to the airspace:

[0095] h g (x,y)=h u (x,y)+h v (x,y);

[0096] Taking the exponents on both sides gives the homomorphic filtering result:

[0097]

[0098] According to the homomorphic filtering algorithm flow, frequency filtering is the key to the processing. Gaussian and exponential filters are often used for homomorphic filtering, and the effect of Gaussian filter processing is slightly better than that of exponential filter. This is because the brightness of the steel wire in the image processed by Gaussian filter is uniform, while the brightness of the image processed by exponential filter is uneven, and the highlight of the steel wire is less suppressed.

[0099] 3-2 Enhancement: Improve the contrast and clarity of the image through histogram equalization, contrast enhancement and other methods; The Retinex image enhancement algorithm can achieve a balance in dynamic range compression, edge enhancement and color constancy, so it can adaptively enhance various types of images.

[0100] The image is considered to be composed of the incident image and the reflected image. The incident light shines on the reflective object, and the reflected light enters the human eye through the reflection of the reflective object. The final image can be expressed by the following formula:

[0101] I(x,y)=L(x,y)*R(x,y);

[0102] Where: I(x,y) represents the image signal observed or received by the camera, L(x,y) represents the illumination component of the ambient light, and R(x,y) represents the reflection component of the target object carrying the image detail information;

[0103] Taking the logarithm of both sides of the equation gives:

[0104] Log(R(x,y))=Log(I(x,y))-Log(L(x,y));

[0105] L(x,y) is approximated by the convolution of I(x,y) and a Gaussian kernel:

[0106] L(x,y)=I(x,y)*G(x,y);

[0107] In the above formula, * represents convolution, G(x,y) represents Gaussian kernel;

[0108] Finally, Log(R(x,y)) is quantized to pixel values ​​in the range of 0-255:

[0109] R(x,y)=(Value-Min) / (Max-Min)*255.

[0110] 4) Perform feature extraction on the preprocessed image to extract the characteristics of the surface damage of the wire rope. The feature extraction includes edge detection, contour extraction and texture analysis;

[0111] 4-1 Edge detection is an image processing technique that aims to identify areas in an image where brightness changes dramatically, which usually correspond to the edges or boundaries of objects. Common edge detection algorithms include the Sobel operator and the Canny operator, which determine the location of the edge by calculating the gradient of the image. In wire rope detection, edge detection can help find the outline of the wire rope and possible damaged areas, such as broken wires, wear, etc.

[0112] 4-2 Contour extraction is to further extract continuous and closed edge lines based on edge detection. It can be implemented through the contour extraction function in the image processing library, such as the findContours function in OpenCV. The extracted contour can be used to more clearly understand the shape and structure of the wire rope, as well as the location and range of damage. In addition, contour extraction can also be used to calculate the diameter, circumference and other parameters of the wire rope, providing a basis for the quality assessment of the wire rope.

[0113] 4-3YOLOv5 target detection is a deep learning-based target detection algorithm that can quickly and accurately identify objects in images. YOLOv5 uses a neural network architecture called Darknet and learns the characteristics of objects through a large amount of training data. In wire rope detection, YOLOv5 can be used to detect damage on the surface of wire ropes, such as broken wires, wear, etc. YOLOv5 can output information such as the location, size, and category of the damage, providing a basis for subsequent processing and analysis.

[0114] 4-4 Texture analysis describes the content of an image by analyzing its texture features, which include gray-level co-occurrence matrix, local binary pattern, etc. In wire rope inspection, texture analysis can help understand the roughness and texture direction of the wire rope surface, so as to determine the type and severity of damage. For example, the surface of a worn wire rope will usually become smoother, while the surface of a broken wire rope will show obvious texture changes.

[0115] 4-5 These methods can be used in combination to obtain more comprehensive and accurate surface damage feature information of wire ropes. For example, edge detection and contour extraction can be used to find the contour of the wire rope, then YOLOv5 can be used to detect the damaged area, and finally texture analysis can be used to further analyze the characteristics of the damage. In this way, the accuracy and reliability of damage detection can be improved, providing protection for the safe operation of wire ropes. It should be noted that the specific implementation and parameter selection of these methods need to be adjusted and optimized according to actual conditions. For example, different types of wire ropes may require different edge detection algorithms and parameters; different types of damage may require different texture features to describe. In addition, the speed and efficiency of image processing need to be considered to meet the requirements of real-time detection.

[0116] 5) Use damage identification algorithm to analyze the extracted features to determine whether there is damage on the surface of the wire rope;

[0117] 5-1 Use video calculation method to obtain the damaged frame time: First select the video, including date, wire rope number and camera position, then perform OpenCV decoding to obtain video frame rate, encoding method, image height and width and other information; then, the system detects each frame of the image, and if wire rope damage is detected, the formula is used Calculate the time of the damaged frame and find the corresponding position in the txt file table. In the formula, k is the frame number of the damaged frame, f2 is the frame rate of the video read, and t2 is the time of the damaged frame.

[0118] 5-2 Removal of repeated damage: After the video calculation results are output, when the first damage is detected, x0 and h0 are recorded, i represents the number of damages, and i is 1 at this time; when the second damage is detected, x1 and h1 are recorded, and then the algorithm enters the formula:

[0119] |(x1-x0)*c+(h1-h0)|>T is used for judgment. If the conditions are met, the damage is saved and the number of damages counted is increased by 1; otherwise, h1 is assigned to h0, x1 is assigned to x0, and then the next detection is performed until all images are detected. In the above formula, c represents the ratio between the pixel size in the image and the actual size, and T represents the error size.

[0120] 6) Output the detection results of wire rope surface damage, including damage type, damage location and damage degree;

[0121] By integrating the detection device and the detection method into one system, real-time online detection of wire rope surface damage is achieved. At the same time, a user interface is developed to facilitate operators to operate and view the detection results. The system and method have the advantages of high efficiency, accuracy, and reliability, and can be widely used in the detection of various wire ropes.

[0122] In addition, the following points should be noted when executing the wire rope surface damage detection method in the embodiment of the present application:

[0123] Real-time optimization: Use multi-threading or parallel processing technology to improve the execution efficiency of the algorithm. Optimize code implementation to reduce unnecessary calculations and memory usage. Algorithm selection and optimization: Select suitable image processing algorithms and deep learning models according to actual needs, and perform targeted optimization and adjustments. System integration and testing: Integrate each module into a complete system, and conduct comprehensive testing and verification to ensure the stability and reliability of the system. Continuous improvement: Keep an eye on the latest technological developments and research results, upgrade and improve the system to improve detection performance and adapt to different application scenarios.

[0124] like Figure 4As shown, the second aspect of the embodiment of the present application provides a wire rope surface damage detection device, which includes a binocular vision detection device, a light source 2, an image acquisition card 3, a 220V power supply 6, an image processing module 5 and a computer 4. The binocular vision detection device includes at least two cameras 1, and the two cameras 1 can simultaneously capture images of the wire rope surface; the light source 2 is used to illuminate the wire rope surface; the image acquisition card 3 is used to capture images captured by the camera; the 220V power supply 6 is used to power the binocular vision detection device, the light source 2 and the computer 4, etc. The image processing module 5 is used to pre-process, extract features and identify damage to the image, and the computer 4 is used to process the image captured by the image acquisition card and output the detection result of the wire rope surface damage.

[0125] Specifically, the camera 1 is an industrial camera, the light source 2 is an LED light source, and the image acquisition card 3 is a high-speed image acquisition card. The wire rope surface damage detection device has the advantages of simple structure, convenient operation, low cost, etc.

[0126] The above are only preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for detecting surface damage of a steel wire rope, characterized in that: The following steps are involved: 1) Arrange a binocular vision detection device, which includes at least two cameras, and adjust the positions of the cameras and the light source so that the surface of the wire rope can be clearly photographed; 2) collecting an image of the surface of the steel wire rope by the camera and transmitting the image to a computer; 3) preprocessing the image to at least improve the quality and contrast of the image; 4) performing feature extraction on the preprocessed image to extract features of the surface damage of the wire rope, wherein the feature extraction includes edge detection, contour extraction and texture analysis; 5) Using a damage identification algorithm to analyze the extracted features to determine whether there is damage on the surface of the wire rope; 6) Outputting the detection result of the surface damage of the steel wire rope, wherein the detection result includes the damage type, damage location and damage degree.

2. The method for detecting surface damage of a steel wire rope according to claim 1, characterized in that: In step 3), the preprocessing of the image includes homomorphic filtering to remove noise in the image, which specifically includes the following steps: Assume that a digital image is g(x, y), its incident component is u(x, y), and its reflected component is v(x, y), then: g(x,y)=u(x,y)·v(x,y); Take the logarithm of both sides: lng(x,y)=lnu(x,y)+lnv(x,y); Take the Fourier transform of both sides: G(u,v)=U(u,v)+V(u,v); Suppose that the filter function H(u, v) is used to process G(u, v): H(u,v)G(u,v)=H(u,v)U(u,v)+H(u,v)V(u,v); Then switch back to the airspace: h g (x,y)=h u (x,y)+h v (x,y); Taking the exponents on both sides gives the homomorphic filtering result:

3. The method for detecting surface damage of a steel wire rope according to claim 1, characterized in that: In step 3), the preprocessing of the image includes binarization processing to adaptively enhance the image, which specifically includes the following steps: The image is considered to be composed of an incident image and a reflected image. The incident light shines on the reflective object, and the reflected light enters the human eye through reflection by the reflective object. The final image can be expressed by the following formula: I(x,y)=L(x,y)*R(x,y); Where: I(x, y) represents the image signal observed or received by the camera, L(x, y) represents the illumination component of the ambient light, and R(x, y) represents the reflection component of the target object carrying the image detail information; Taking the logarithm of both sides of the equation gives: Log(R(x,y))=Log(I(x,y))-Log(L(x,y)); L(x, y) is approximated by the convolution of I(x, y) and a Gaussian kernel: L(x,y)=I(x,y)*G(x,y); In the above formula, * represents convolution, G(x, y) represents Gaussian kernel; Finally, Log(R(x, y)) is quantized to pixel values ​​in the range 0-255: R(x,y)=(Value-Min) / (Max-Min)*255.

4. The method for detecting surface damage of a steel wire rope according to any one of claims 1 to 3, characterized in that: In the step 2), when collecting the image of the wire rope surface, it is necessary to clarify the relationship between the camera parameters and the wire rope parameters, including the relationship between the parameters when collecting the full rope information, the relationship between the actual field of view and the camera imaging surface, and the relationship between the camera parameters and the wire rope speed when running at high speed; Among them, the relationship between the running speed of the wire rope, the running time, the camera frame rate, the actual field of view height, and the overlap of adjacent images is: Fh-(F-1)s=vt; where F is the camera frame rate, s is the overlap between two adjacent images, v is the running speed of the wire rope, t is the running time of the wire rope, and h is the actual field of view height of the wire rope; The basic parameters of the camera and lens: S = h × l, so: F (h × l × D) - f (F-1) s = vtf; where W is the actual optical magnification, f is the focal length, D is the object distance, S is the imaging height, h resolution is the height of the image resolution, l pixel is the size of the camera pixel, and h is the actual field of view height of the wire rope.

5. The method for detecting surface damage of a steel wire rope according to claim 4, characterized in that: In the step 2), when performing image acquisition on the surface of the steel wire rope, the acquisition strategy used is to establish a camera class in the algorithm, and coordinate resources on the basis of multi-camera instantiation through a multi-threaded parallel acquisition structure; The data storage strategy used is as follows: after the system and program are started, the algorithm starts three threads to process the data and position data of the two cameras respectively, and the computer saves while collecting; when storing, a cache area is first opened in the memory, which consists of at least 4 one-dimensional arrays, each of which can store 300 frames of images, occupying 9G of memory and 32G of system memory; when a cache area is filled, the save flag flag+1 is saved, the saving thread starts saving from the first array, and the acquisition thread continues to collect and save to the first array, and so on; When the acquisition thread stops, the saving thread will continue to save until the images in the cache area are emptied, and then use avi video encoding technology to encode the images to form a video, and finally save the video and location data to the hard disk and txt file respectively.

6. The method for detecting surface damage of a steel wire rope according to any one of claims 1 to 3 or 5, characterized in that: In the step 4), edge detection algorithm is used to extract edge features of the surface of the steel wire rope, and the edge detection algorithm determines the position of the edge by calculating the gradient of the image, so as to find the outline of the steel wire rope and possible damaged areas; The contour extraction is performed by using a contour extraction function in an image processing library, wherein the contour extraction function includes a findContours function in OpenCV; When performing the texture analysis, the content of the image is described by analyzing the texture features of the image, and the texture features include gray-level co-occurrence matrix and local binary pattern.

7. The method for detecting surface damage of a steel wire rope according to any one of claims 1 to 3 or 5, characterized in that: In the step 5), the support vector machine algorithm is used to classify the extracted features to determine whether there is damage on the surface of the wire rope. The extracted features are analyzed including: The video calculation method is used to obtain the damaged frame time: first select the video, including the date, wire rope number and camera position, and then perform OpenCV decoding to obtain the video frame rate, encoding method, image height and width information; then, the system detects each frame of the image, and if the wire rope is damaged, the formula is used to calculate the damage frame time. Calculate the time of the damaged frame and find the corresponding position in the txt file table. In the formula, k is the frame number of the damaged frame, f2 is the frame rate of the video read, and t2 is the time of the damaged frame. Removal of repeated damage: After the video calculation results are output, when the first damage is detected, x0 and h0 are recorded, i represents the number of damages, and i is 1 at this time; when the second damage is detected, x1 and h1 are recorded, and then the algorithm enters the formula: |(x1-x0)*c+(h1-h0)|>T, and makes a judgment. If the conditions are met, the damage is saved, and i, which counts the number of damages, is added by 1; otherwise, h1 is assigned to h0, x1 is assigned to x0, and then the next detection is performed until all images are detected; in the above formula, c represents the ratio between the pixel size in the image and the actual size, and T represents the error size.

8. The method for detecting surface damage of a steel wire rope according to any one of claims 1 to 3 or 5, characterized in that: In the step 1), adjusting the camera includes adjusting the camera parameters, specifically including: Adjust the camera's focus, aperture, shutter speed, and adjust camera parameters according to specific environment and needs; Initialize the parameters, including setting the camera frame rate to 100 frames to ensure that sufficient image details can be captured, setting the resolution to 1280×1024 to obtain a clear image, and setting the saved video format and file address for subsequent storage and processing of image data; Start the camera and open three threads, including two camera threads and one position data acquisition thread. The two camera threads are used to collect image data from two cameras at the same time to realize the binocular vision function. The position data acquisition thread is used to obtain the camera's position information for subsequent image processing and analysis.

9. A wire rope surface damage detection device, characterized in that: include: A binocular vision detection device, the binocular vision detection device comprising at least two cameras, the two cameras being capable of simultaneously capturing images of the surface of the wire rope; A light source, the light source is used to illuminate the surface of the wire rope; An image acquisition card, wherein the image acquisition card is used to acquire images taken by the camera; An image processing module and a computer, wherein the image processing module is used for preprocessing, feature extraction and damage identification of images, and the computer is used for processing the images collected by the image acquisition card and outputting the detection results of the surface damage of the wire rope.

10. The wire rope surface damage detection device according to claim 9, characterized in that: The camera is an industrial camera, the light source is an LED light source, and the image acquisition card is a high-speed image acquisition card.

Images