A detection method for spacer bar vibration test results based on machine vision

Abstract

The present invention relates to the field of image data processing technology, and in particular to a spacer bar vibration test result detection method based on machine vision, comprising: collecting images of rigid spacers before and after the vibration test, and obtaining a target image based on the images; the target image is: using a set sliding window to traverse the image, calculating the variance of the grayscale values ​​of all pixels in the set sliding window, thereby adjusting the weight of each pixel, and combining the grayscale and gradient intensity information to calculate the grayscale value of each pixel in the target image; generating a target image based on the grayscale value of each pixel, and in response to the fact that the slip amount of the center of gravity position of the wire clamp of the rigid spacer in the target image before and after the vibration test is within a preset slip range, it is determined that the fixity of the rigid spacer is normal. The present invention solves the problems of noise interference and inaccurate feature extraction when performing image data processing.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and more specifically, to a method for detecting spacer bar vibration test results based on machine vision. Background Art

[0002] With the development of the power system and the continuous expansion of the power grid, the safe operation of transmission lines is becoming increasingly important. As a key component in high-voltage transmission lines, spacers are mainly used to maintain a reasonable spacing between conductors and reduce damage to the conductors caused by external factors such as wind and vibration. However, spacers are exposed to the outdoor environment for a long time and are often affected by factors such as wind, rain, temperature changes, and electric fields. They are prone to vibration and fatigue damage, which in turn affects the reliability and stability of the transmission lines. In order to ensure the normal working state of the spacers, regular inspections have become an important part of power equipment maintenance.

[0003] At present, the spacer detection technology based on machine vision determines the vibration state and damage of the spacer through image acquisition, feature extraction and anomaly detection, and extracts key feature parameters using grayscale transformation, noise suppression, edge detection and other processing methods to analyze its surface state. However, due to factors such as light and weather in complex detection environments, images are easily disturbed by noise, resulting in unstable processing results; in addition, existing feature extraction methods are difficult to capture subtle changes caused by vibration, which can easily lead to misjudgment or missed judgment, affecting the reliability of detection results.

[0004] The patent application document with application publication number CN104933403A discloses a method for monitoring conductor galloping based on spacer bar identification. The patent application document takes the digital image intercepted from the video stream of the transmission line as the research object, and uses pre-processing methods such as graying and smoothing denoising to perform threshold segmentation, edge detection and contour tracking on the digital image to extract the contour of the conductor spacer bar. By comparing the template contours before and after the galloping, the galloping torsion angle is calculated, and when an abnormal angle is detected, an early warning message is sent to the safety inspector in a timely manner, so that the detection and early warning of the galloping torsion angle of the spacer bar conductor can be realized quickly and automatically.

[0005] However, although the above scheme has achieved certain results in detecting anomalies of spacer bars, there are still problems of noise interference and inaccurate feature extraction when processing image data, which may lead to deviations in the detection results and affect the accuracy of the final detection. Summary of the invention

[0006] In order to solve the problems of noise interference and inaccurate feature extraction in image data processing raised in the above background technology, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a method for detecting the results of a spacer vibration test based on machine vision, comprising: collecting images of a rigid spacer before and after a vibration test, and acquiring a corresponding target image based on each image; wherein the gray value of each pixel in the target image is:

[0008] ;

[0009] In the formula, The target image Line The gray value of the pixel. For the image Line The gray level to which the column pixel belongs, For the image Line The gradient strength of the column pixel, For the Line Column pixel weight; weight for:

[0010] ;

[0011] In the formula, is a parameter that controls the steepness of the function. , The image is Line The variance of the grayscale values ​​of all pixels in the set sliding window centered on the column pixel point is calculated; if the slip amount of the center of gravity position of the wire clamp of the rigid spacer in the target image before and after the vibration test is within the preset slip range, the fixity of the rigid spacer is judged to be normal.

[0012] The above technical solution collects the images of the rigid spacer bar before and after the vibration test, and generates the target image by weighted synthesis method, combines the gray value and gradient intensity of the pixel point, and smoothes each pixel based on the adaptive weight coefficient, so that the target image can more clearly present the subtle differences between the images before and after the vibration; the calculation of the weight takes into account the local grayscale variance, improves the response sensitivity to the local features of the image, and thus ensures the effective expression of the image details. At the same time, by calculating the slippage of the center of gravity position of the rigid spacer bar wire clamp and comparing it with the preset slippage range, it is possible to accurately determine whether the fixity of the spacer bar is normal, ensuring the reliability and accuracy of the determination process.

[0013] Furthermore, the center of gravity position of the wire clamp is: using the Canny edge detection algorithm to extract the outline of the rigid spacer rod wire clamp in the target image before and after the vibration test, and obtaining the minimum circumscribed rectangle of the outline, and the center point position of the minimum circumscribed rectangle is the center of gravity position of the wire clamp.

[0014] The above technical solution accurately extracts the contour of the rigid spacer wire clamp in the target image before and after the vibration test, ensuring the accurate capture of key edge features. Then, by calculating the minimum circumscribed rectangle of the contour and determining the position of its center point, the center of gravity of the wire clamp is accurately located. This method improves the stability and consistency of the center of gravity positioning, making the tracking of the center of gravity position of the wire clamp under different vibration states more reliable, thereby effectively supporting the determination of the fixity of the rigid spacer and enhancing the accuracy and reliability of the overall detection process.

[0015] Furthermore, it also includes detecting the vibration reduction performance of the rigid spacer bar before and after the vibration test: using the Canny edge detection algorithm to extract the area where the rigid spacer bar rubber element is located in the target image before and after the vibration test, and marking the number of connected domains in the area where the rigid spacer bar rubber element is located in the target image before and after the vibration test; if the change in the number of connected domains in the area where the rigid spacer bar rubber element is located in the target image before and after the vibration test is within a preset number range, it is determined that the vibration reduction performance of the rigid spacer bar is normal.

[0016] The above technical solution accurately extracts the area where the rigid spacer rubber element is located in the image before and after the vibration test, and combines the connected domain analysis method to calculate the number of connected domains in the rubber element area. By comparing the changes in the number of connected domains of the rubber element before and after the test and setting a reasonable range of changes, it is possible to effectively determine whether the rubber element has undergone significant deformation or wear, thereby indirectly evaluating whether the vibration reduction performance of the rigid spacer is normal. This solution improves the sensitivity and accuracy of vibration reduction performance detection, and can identify potential wear problems of rubber elements earlier, thereby providing strong protection for the health of the equipment.

[0017] Furthermore, the grayscale is 8 bits and the range is [0, 255].

[0018] Furthermore, a CMOS camera or a CCD camera is used to collect images of the rigid spacer before and after the vibration test.

[0019] Furthermore, the method also includes performing image grayscale conversion and median filtering on the image.

[0020] The above technical solution effectively simplifies the color information of the image, converts the image into a single grayscale channel, reduces data complexity, and facilitates subsequent analysis. At the same time, median filtering is used to remove noise in the image, especially for random noise that may be generated in vibration tests. Median filtering can effectively retain edge details and avoid blurring key features. This preprocessing step makes the image clearer and smoother, providing higher quality input data for subsequent edge detection and feature extraction, and improving overall detection accuracy and stability.

[0021] Furthermore, the Sobel operator algorithm is used to obtain the gradient intensity of each pixel in the image before and after the rigid spacer vibration test.

[0022] The above technical solution effectively highlights the edge and structural details in the image, especially the edge features of the rigid spacer. The use of gradient intensity can more comprehensively reflect the overall and local change trends of the image, thereby enhancing the sensitivity to subtle deformation and position changes before and after the vibration test. It provides rich data information for further analysis of the stability of the rigid spacer during vibration, helps to accurately evaluate the structural changes under the influence of the test, and improves the accuracy and reliability of detection.

[0023] Furthermore, a histogram equalization algorithm is used to obtain the gray level of each pixel in the image before and after the rigid spacer vibration test.

[0024] The above technical solution uses a histogram equalization algorithm to adjust the grayscale of each pixel in the image before and after the rigid spacer vibration test, which can effectively improve the contrast and detail performance of the image, especially in low-contrast areas, which helps to enhance subtle changes and detail features in the image. During the vibration test, the rigid spacer may undergo slight deformation or position changes. Histogram equalization can optimize the overall visual effect of the image, making these changes easier to detect and analyze. In addition, the equalization process can also reduce the impact of uneven lighting or shadows, improve the overall quality of the image, and thus provide more reliable and accurate data support for subsequent image analysis, feature extraction and detection algorithms, and enhance the sensitivity and accuracy of the system.

[0025] The beneficial effects of the present invention are:

[0026] The present invention detects the fixity and vibration reduction performance of the spacer before and after the vibration test by a method based on machine vision, and adopts a variety of image processing technologies to improve the accuracy and stability of the detection. By obtaining the gray value and gradient intensity of each pixel in the image, and combining weighted processing and adaptive parameters, the slight changes in the image before and after the test can be presented more clearly. The contours of the key components of the spacer and the number of connected domains in the rubber element area are extracted using Canny edge detection to intuitively reflect the displacement and wear after vibration. Combined with image graying, median filtering, histogram equalization and Sobel operator and other processing methods, the clarity of the image and the noise removal effect are further improved, making the detection results more accurate and reliable, and significantly improving the performance detection effect of the spacer based on machine vision. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0028] Figure 1 is a flow chart schematically showing a method for detecting a spacer bar vibration test result based on machine vision according to an embodiment of the present invention;

[0029] Figure 2 is a grayscale image schematically illustrating a rigid spacer according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0031] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0032] An embodiment of a method for detecting spacer bar vibration test results based on machine vision.

[0033] like Figure 1 As shown, a flow chart of a spacer bar vibration test result detection method based on machine vision according to an embodiment of the present invention comprises the following steps:

[0034] S1: Images of the rigid spacer bar before and after the vibration test.

[0035] In one embodiment, a high-resolution CMOS camera or CCD camera can be used to capture images of the state of the rigid spacer before and after the vibration test to ensure that the details of the image are clearly visible. Among them, the CMOS camera has the advantages of low power consumption and fast response, and can accurately capture the subtle changes of the rigid spacer during the vibration process; while the CCD camera has high imaging quality and light sensitivity, and can obtain high-quality images in low-light environments. By selecting the appropriate camera type, the accuracy and stability of the image can be effectively improved, thereby providing a high-quality data foundation for subsequent image processing.

[0036] In another embodiment, other types of cameras may be used for image acquisition, such as infrared cameras, etc., to adapt to different environmental conditions and test requirements.

[0037] In order to further improve the image quality, the collected original images usually need to be preprocessed, because the collected original images usually contain noise and uneven light distribution. Therefore, the image needs to be grayed first, and the color image is converted into a gray image, so as to reduce the data dimension and facilitate subsequent processing. On the basis of the grayed image, median filtering is further performed to remove salt and pepper noise and other local noise. As a nonlinear filtering method, median filtering can effectively retain edge information while reducing image noise, thereby obtaining a smoother image. This series of image processing steps significantly improves the image quality. After graying and median filtering, the image has effectively suppressed noise interference, and the edges and details are more prominent. The stability and reliability of the image data have been greatly enhanced, providing a clearer and more accurate basis for subsequent image analysis. Through these optimization processes, it can be ensured that the image can more accurately reflect the subtle changes of the rigid spacer in the vibration test in the subsequent edge detection, feature extraction and state comparison analysis, thereby supporting efficient and reliable performance detection and evaluation.

[0038] S2: and obtaining a corresponding target image based on each image.

[0039] In one embodiment, the grayscale value of each pixel in the target image is:

[0040] ;

[0041] In the formula, The target image Line The gray value of the pixel. For the image Line The gray level to which the column pixel belongs, For the image Line The gradient strength of the column pixel, For the Line Column pixel weight; weight for:

[0042] ;

[0043] In the formula, is a parameter that controls the steepness of the function. , The image is Line The variance of the grayscale values ​​of all pixels in the set sliding window centered on the column pixel.

[0044] In one embodiment, the size of the sliding window can be 5 5. Of course, it can also be set based on experience. The variance of the sliding window represents the grayscale fluctuation of the pixels in the window, which can reflect the richness of the edges in the area. A high variance value indicates that the grayscale changes in the window are significant and may contain more edges, while a low variance value corresponds to fewer edges. The weight is controlled by the variance in the window and is adjusted by the parameter Adjustment is made. As the variance increases, the weight becomes larger, which means that the grayscale of the original pixel will be more emphasized in the multi-edge area, thereby improving the detail performance of the image; conversely, for the area with fewer edges, the weight will be more biased towards the gradient intensity to reduce the impact of noise.

[0045] The grayscale is 8 bits and the range is [0,255]. By limiting the grayscale to 8 bits, the storage and processing of the image can be effectively simplified, while ensuring the color accuracy and detail information of the image, and ensuring sufficient dynamic range during the processing. An 8-bit grayscale image can provide 256 grayscale levels, which provides a good information basis for subsequent image analysis, especially when performing subtle change detection, it can accurately reflect the grayscale differences and structural changes of the image.

[0046] In addition, the scheme uses the Sobel operator algorithm to calculate the gradient intensity of each pixel in the image. The Sobel operator is a commonly used edge detection algorithm that extracts the edge information of the image by calculating the gradient of the image in the horizontal and vertical directions. Specifically, the Sobel operator can highlight the changing areas in the image, especially the edge areas, by applying a gradient filter to the image. In the vibration test image, this edge detection helps to clearly identify the deformation and structural changes that may occur in the rigid spacer during the vibration process, such as the slippage of the wire clamp position or the deformation of the rubber element. Therefore, the use of the Sobel operator can effectively extract the edge information in the image, providing a solid foundation for subsequent target detection and structural analysis.

[0047] On this basis, the histogram equalization algorithm is further applied. This algorithm enhances the contrast of the image by adjusting the grayscale distribution of the image, especially in low-contrast images. Histogram equalization redistributes the grayscale of the image to make the brightness information in the image more uniform and highlight the important structural details in the image. For example, in the images before and after the vibration test, by enhancing the grayscale contrast, the tiny vibration deformation or other details are more significant, so that the deformation and damage parts of the rigid spacer can be more accurately identified. This contrast enhancement processing method helps to improve the accuracy of image analysis, especially in images with low contrast or complex backgrounds, which can effectively improve image quality and enhance the robustness of the detection algorithm.

[0048] The above scheme controls the weights through the variance of the sliding window, which can enhance the grayscale ratio of the original pixels in the area with many edges, making the details of the image clearer; while in the area with few edges, the scheme will suppress the grayscale noise and pay more attention to the gradient intensity, thereby reducing noise interference and improving the overall smoothness of the image.

[0049] S3: In response to the fact that the amount of slippage of the center of gravity position of the wire clamp of the rigid spacer in the target image before and after the vibration test is within a preset slip range, it is determined that the fixation of the rigid spacer is normal.

[0050] The setting of the preset slip range is usually adjusted according to the accuracy requirements of the vibration test, the bearing capacity of the equipment, the material properties and the structural design. Generally speaking, the slip range will take into account the accuracy tolerance of the equipment, the material and structural characteristics of the spacer bar, the test purpose and the safety standard. In this embodiment, the preset slip range is between 0.1mm and 2mm. However, the specific slip range needs to be appropriately adjusted according to the vibration intensity, the working environment and the working requirements of the equipment to ensure that it can meet the performance verification without affecting the reliability and safety of the equipment.

[0051] In one embodiment, the center of gravity of the clamp is:

[0052] The Canny edge detection algorithm is used to extract the outline of the rigid spacer wire clamp in the target image before and after the vibration test, and the minimum circumscribed rectangle of the outline is obtained. The center point position of the minimum circumscribed rectangle is the center of gravity position of the wire clamp. The above method effectively identifies the outline of the wire clamp through the Canny edge detection algorithm, and simplifies the calculation of the center of gravity by using the geometric features of the minimum circumscribed rectangle, avoiding complex shape analysis and improving processing speed and accuracy. Using the center point of the minimum circumscribed rectangle as the center of gravity position can quickly reflect the displacement change of the wire clamp during the vibration test, adapt to the fixity detection requirements of most standard-shaped wire clamps, and provide strong data support for subsequent fault diagnosis and performance analysis.

[0053] In another embodiment, the center of gravity position of the wire clamp can be determined by calculating the center of mass of the wire clamp area. First, the wire clamp area is extracted using an image segmentation method, and then the weighted average coordinates of all pixel points in the area are calculated using a weighted average algorithm to obtain the center of mass position of the wire clamp area. This method does not rely on the geometry of the circumscribed rectangle, and can also more accurately reflect the actual center of gravity of the wire clamp, and is particularly suitable for wire clamps with complex or irregular shapes. The center of mass calculation can improve positioning accuracy and adapt to image processing requirements in more different situations.

[0054] In one embodiment, the vibration reduction performance of the rigid spacer bar in the target image before and after the vibration test can be detected by the following steps: First, the contour of the rigid spacer bar rubber element in the image before and after the vibration test is accurately extracted using the Canny edge detection algorithm to ensure that the area where the rubber element is located is effectively identified. Then, the connected domain marking algorithm is used to analyze the extracted rubber element area and mark the number of connected domains of each rubber element area in the image. Finally, the change in the number of connected domains in the rubber element area in the image before and after the vibration test is compared. If the change value is within a preset number range, it is determined that the vibration reduction performance of the rigid spacer bar is normal. Through this method, the performance change of the rubber element in the vibration test can be accurately judged, especially the high sensitivity to the slight change in the vibration reduction effect, which helps to discover potential vibration reduction performance degradation or failure problems at an early stage, thereby improving the reliability and safety of the rigid spacer bar.

[0055] In another embodiment, the vibration reduction performance of the rigid spacer in the target image before and after the vibration test can also be detected by analyzing the deformation of the rubber element in the image. First, image preprocessing methods such as graying and filtering are used to remove noise and improve image quality; then, the contour of the rubber element in the image before and after the vibration test is extracted by edge detection algorithms such as Canny or Sobel operators, and the change of the contour, such as morphological features: contour area, perimeter, etc., is calculated; then, by comparing the geometric changes of the rubber element before and after the test, it is determined whether its deformation amplitude meets the preset standard, thereby determining whether the vibration reduction performance is normal. This method can detect the slight deformation of the rubber element during the vibration process, thereby more accurately evaluating its vibration reduction effect, helping to detect the decline in vibration reduction performance in advance, and ensuring the safety and reliability of the rigid spacer.

[0056] like Figure 2 , which is a grayscale image of the rigid spacer bar according to an embodiment of the present invention.

[0057] The scheme of the present invention realizes high-precision detection of the state of the rigid spacer before and after the vibration test through a multi-step image processing method, ensuring reliable evaluation of fixity and vibration reduction performance. First, the target image generation method based on grayscale value and gradient intensity enhances the detail presentation of the image through adaptive weights, which is convenient for accurately identifying the displacement of the rigid spacer. Secondly, the combination of the Canny edge detection algorithm and the minimum circumscribed rectangle accurately locates the center of gravity slip of the wire clamp, which is convenient for stability analysis. In addition, the connected domain analysis method is used to evaluate the connectivity changes of the rubber element, which improves the sensitivity of vibration reduction performance detection. The application of various image processing techniques such as graying, median filtering, Sobel operator and histogram equalization makes the image noise suppression, gradient enhancement and contrast enhancement effects significant, greatly improves the quality of image data and the accuracy of detection, and provides a reliable image analysis basis for the scientific evaluation of the performance of rigid spacer bars.

[0058] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0059] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for detecting spacer bar vibration test results based on machine vision, characterized in that: include: Collecting original images of the rigid spacer bar before and after the vibration test, and acquiring the first image of the corresponding rigid spacer bar before and after the vibration test based on each original image, so that the noise interference of the original image is effectively suppressed, and the edge and details are more prominent; Among them, the grayscale value of each pixel in the first image is: ； In the formula, The first image Line The gray value of the pixel. is the first Line The gray level to which the column pixel belongs, is the first Line The gradient strength of the column pixel, For the Line Column pixel weights; Weight for: ； In the formula, is a parameter that controls the steepness of the function. , is the original image with Line The variance of the grayscale values ​​of all pixels in the set sliding window centered at the column pixel; If the amount of slip of the center of gravity position of the wire clamp of the rigid spacer bar in the first image before and after the vibration test is within a preset slip range, it is determined that the fixation of the rigid spacer bar is normal.

2. The spacer bar vibration test result detection method based on machine vision according to claim 1 is characterized in that: The center of gravity position of the clamp is: The Canny edge detection algorithm is used to extract the outline of the rigid spacer wire clamp in the first image before and after the vibration test, and the minimum circumscribed rectangle of the outline is obtained. The center point position of the minimum circumscribed rectangle is the center of gravity position of the wire clamp.

3. The method for detecting the spacer bar vibration test results based on machine vision according to claim 1, characterized in that: It also includes testing the vibration damping performance of the rigid spacer bars before and after the vibration test: The Canny edge detection algorithm is used to extract the area where the rigid spacer rod rubber element is located in the first image before and after the vibration test, and the number of connected domains in the area where the rigid spacer rod rubber element is located in the first image before and after the vibration test is marked respectively; If the change in the number of connected domains in the region where the rigid spacer rubber element is located in the first image before and after the vibration test is within a preset number range, it is determined that the vibration reduction performance of the rigid spacer is normal.

4. The method for detecting the spacer bar vibration test results based on machine vision according to claim 1, characterized in that: The grayscale is 8 bits and ranges from [0, 255].

5. The spacer rod vibration test result detection method based on machine vision according to claim 1, characterized in that: The original images of the rigid spacer before and after the vibration test are collected using a CMOS camera or a CCD camera.

6. The method for detecting the spacer bar vibration test result based on machine vision according to claim 1, characterized in that: The method also includes performing image grayscale conversion and median filtering on the original image.

7. The method for detecting the spacer bar vibration test result based on machine vision according to claim 1, characterized in that: The Sobel operator algorithm is used to obtain the gradient intensity of each pixel in the original image before and after the rigid spacer vibration test.

8. The method for detecting spacer bar vibration test results based on machine vision according to claim 1, characterized in that: The histogram equalization algorithm is used to obtain the gray level of each pixel in the original image before and after the rigid spacer vibration test.

Images