Cable force testing method based on infrared light and optimal self-adaptive template matching

Abstract

A kind of cable force test method based on infrared light and optimal adaptive template matching, characterized by, including collecting cable vibration video;Image preprocessing;Select template picture and search area;Sub-pixel processing;Target tracking;Spectrum analysis;Cable force identification.The present application solves the problem that the existing machine vision technology is difficult to test cable force in bad weather conditions, and the accuracy is poor, and it is difficult to test cable force in weak light intensity environment;And image noise reduction algorithm is used to reduce image noise, then bilinear interpolation algorithm is used to improve the resolution of image, optimal adaptive template matching algorithm is used for target tracking, so that the displacement time history curve of cable is obtained, the natural frequency of cable is obtained by using fast fourier transform to analyze the spectrum of displacement time history data, and the cable force is calculated according to the cable force-frequency formula.

Description

Technical Field

[0001] This invention belongs to the field of bridge health monitoring technology, specifically relating to a cable tension testing method based on infrared light and optimal adaptive template matching. Background Technology

[0002] In recent decades, with the rapid development of the social economy and the advancement of urbanization, the transportation industry has flourished, and more and more bridges have been built to meet the growing demand for freight transport and travel. Among them, cable-stayed bridges, suspension bridges, and arch bridges are widely used in long-span bridges due to their aesthetic appeal, cost-effectiveness, and strong spanning capacity. Cables, as the main load-bearing components of cable-stayed bridges, are typically made of high-strength steel wire and strand, such as the stay cables in cable-stayed bridges, the suspenders and main cables in suspension bridges, and the suspenders and tie rods in arch bridges. Cable-stayed bridges are generally highly statically indeterminate structures, and the distribution of internal forces largely depends on the distribution of cable forces. During use, cables are prone to stress relaxation, prestress loss, corrosion, and other defects, leading to a redistribution of internal forces. Bridge collapses caused by cable damage are frequent. Therefore, cable force monitoring is of great significance to the safety of cable-stayed bridges, and it is necessary to conduct cable force monitoring during the construction and operation phases of the bridge.

[0003] Currently, commonly used cable force testing methods include hydraulic pressure gauge measurement, pressure sensor measurement, frequency method, magnetic flux method, and machine vision method. Among these, machine vision-based methods have advantages such as visualization, low cost, convenient installation, and wide applicability, and have been widely used. For example, patents such as CN114972295A (Non-contact cable force recognition technology based on multi-point displacement extraction), CN108955983A (Cable force testing method based on cable vibration mode and photogrammetry), CN112381190A (A cable force testing method based on mobile phone image recognition), CN113607321A (Cable force testing method, device, computer equipment, and storage medium for cable structures), and CN111259770A (A rapid cable force testing system and method based on UAV platform and deep learning in complex backgrounds) all employ machine vision methods for cable force testing. While these methods can perform cable force testing under certain conditions, they are difficult and inaccurate under adverse weather conditions (such as rain, fog, and snow). Furthermore, machine vision-based cable force testing methods rely on visual devices for video capture, making it difficult to conduct effective cable force testing in environments with low light intensity (such as nighttime or dark environments).

[0004] Therefore, it is necessary to propose a suitable cable force testing method to solve the problems of difficulty and poor accuracy in cable force testing under adverse weather conditions (such as rain, fog, snow, etc.), and to enable accurate cable force testing in environments with low light intensity (such as nighttime, no light source, etc.), thereby improving measurement efficiency and accuracy.

[0005] Existing machine vision-based methods are all non-contact cable force testing methods. The general process involves installing a target on the cable, using a vision device to capture video of the cable vibration, capturing the displacement-time history curve of the target, performing spectral analysis on the displacement-time history data, and finally calculating the cable force using the frequency-force formula. The drawback of existing technologies is that image acquisition of the target on the cable is greatly affected by external environmental interference. For example, adverse weather conditions (such as rain, fog, and snow) can obscure the target to some extent, making it difficult to continuously acquire clear video of the target vibration, thus significantly increasing the difficulty of cable force testing. Furthermore, video acquisition using vision devices is typically performed under sufficient lighting conditions. However, in low-light conditions (such as at night or in the absence of a light source), the visibility of the cable and target is low, making it impossible to acquire clear video of cable vibration, thus hindering cable force testing. Simultaneously, existing machine vision algorithms have limited recognition performance in low-light, high-frequency, fast-moving scenarios. Summary of the Invention

[0006] To overcome the shortcomings of existing cable force testing technologies, this invention provides a cable force testing method based on infrared light and optimal adaptive template matching. This method enables non-contact cable force testing under adverse weather conditions and low light intensity, improving measurement efficiency and accuracy while ensuring measurement accuracy. The technical solution is as follows: A cable tension testing method based on infrared light and optimal adaptive template matching, characterized by the following steps: Step 1: Collect cable vibration video: Install infrared light targets at the measuring points on the cable, and then use a camera equipped with an infrared filter to collect the vibration video of the cable. Step 2: Image preprocessing: For the acquired cable vibration video, the video frames are first converted into grayscale images using an image preprocessing algorithm, and then the images are denoised. Step 3: Select template image and search area: In the first frame of the video, select the template image and its search area respectively; Step 4: Subpixel processing: Perform bilinear interpolation on the template image and the search region to obtain a higher resolution subpixel-level image; Step 5: Target tracking: Use the optimal adaptive template matching algorithm to track the position of the target in the video and obtain its displacement time history data; Step 6: Spectrum analysis: Perform a fast Fourier transform on the displacement time history data to obtain the natural frequencies of the cable.

[0007] Step 7: Cable Force Identification: Substitute the obtained cable natural frequency into the frequency-cable force formula to calculate the cable force. Beneficial effects

[0008] (1) The present invention realizes non-contact cable force testing based on infrared light and optimal adaptive template matching algorithm. Compared with the traditional contact cable force testing method, which requires the deployment of a large number of sensors or cable force measuring devices, the present invention significantly reduces costs and has higher measurement efficiency and accuracy.

[0009] (2) The present invention uses an infrared light-based image acquisition device, which has stronger penetration than traditional cameras and can capture clearer images under adverse weather conditions and low light intensity. It can solve the problems of high difficulty and poor accuracy of existing machine vision methods in cable force testing under adverse weather conditions (such as rain, fog, snow, etc.) and the difficulty in cable force testing in low light intensity environments (such as night, dark environments, etc.).

[0010] (3) The present invention uses the optimal adaptive template matching algorithm to calculate the displacement time history curve of the cable. When matching the template, the optimal template among the fixed template and the dynamic template is used to improve the matching accuracy, thereby obtaining more accurate displacement time history data, and thus obtaining more accurate natural frequency and cable force.

[0011] (4) The present invention uses a bilinear subpixel interpolation algorithm to obtain a higher resolution template image and search area, and can obtain subpixel-level matching accuracy when matching templates, thereby improving the template matching accuracy. Attached Figure Description

[0012] Figure 1 This invention relates to an infrared light-based image acquisition device; Figure 2 This is a flowchart of the algorithm of the present invention; Figure 3 This is a schematic diagram of an embodiment of the present invention; Figure 4 This is a schematic diagram of bilinear interpolation in this invention. Figure 5 This is the displacement time history diagram of the present invention; Figure 6 This is the spectrum diagram of the present invention. Detailed Implementation

[0013] Existing machine vision methods for testing cable tension in cable-stayed bridges suffer from difficulties and inaccuracies under adverse weather conditions (such as rain, fog, and snow) and in low-light environments (such as nighttime or darkness). This invention utilizes the strong penetrating power of infrared light under adverse weather conditions to propose a cable tension testing method based on matching infrared light with an optimal adaptive template, effectively addressing the shortcomings of existing technologies.

[0014] The flowchart of the optimal adaptive template matching algorithm proposed in this invention is as follows: Figure 2 As shown: Step 1: Collect video of cable vibration: such as Figure 3 As shown, the infrared light target is installed at the measurement point on the cable. The measurement point should be as close as possible to the middle of the cable and unobstructed to facilitate measurement. Then, a camera with an infrared filter is set up in a suitable position. The background of the shooting area should be as simple as possible, and all measurement points should be included in the shooting area. Then, the vibration video of the cable is collected.

[0015] The data acquisition device mainly includes: a camera, an infrared filter, and an infrared target light, such as... Figure 1 As shown, the system includes: a camera, a high-pixel, high-resolution image acquisition device primarily used to capture video of cable vibration; an infrared filter, a camera lens adapted to the infrared band, capable of filtering visible light while allowing infrared light to pass through. Cameras equipped with infrared filters can filter out ambient light interference when acquiring video, capturing only infrared light; and infrared light targets, which emit infrared light. Due to the longer wavelength of infrared light, it has strong penetrating power even in adverse weather conditions (such as rain, fog, snow) and low light intensity conditions (such as nighttime, no light source, etc.). Specifically: A certain number of infrared light targets are installed at measurement points on the cable according to actual needs. Then, the infrared filter is installed on the camera, and the infrared lights are turned on. The camera is then used to capture video of the cable vibration.

[0016] Step 2: Image Preprocessing: Import the acquired cable vibration video into the computer. Use image preprocessing algorithms to convert the RGB three-channel video frame images into single-channel grayscale images to speed up subsequent calculations. Based on this, perform noise reduction processing to decrease image noise. The specific implementation of converting to grayscale is as follows:

[0017] in This indicates that the x-coordinate of the original color image is... The vertical axis is The pixels; Represents pixels in a grayscale image grayscale value, Represents pixels in the original color image The pixel value of the red channel, Represents pixels in the original color image The pixel value of the green channel, Represents pixels in the original color image The pixel value of the blue channel.

[0018] Step 3: Select Template Image and Search Area: In the first frame of the video, select the template image and its search area. The video frames following Step 2 will contain the cable and at least one infrared target light at the measurement point location. The measurement point number is... .like Figure 3 As shown, there are three measurement points, and an infrared target is installed at each measurement point. Then, in the first frame of the video, the infrared target at each measurement point is selected as a template image. The requirement is that the selected area should exactly encompass the template image, as shown below. Figure 3 As shown, the vibration range of the measuring points on the cable is analyzed. Based on the movement range of the measuring points in the video, and with a certain safe movement range reserved, a search area for each measuring point is defined. This search area is then selected as the search area for each frame in subsequent template matching. The size of the search area should be sufficient to encompass the movement range of the measuring points, and the search area for each measuring point should not be too large to reduce the time required for subsequent template matching.

[0019] Step 4: Subpixel Processing: Since image resolution has a certain impact on the accuracy of subsequent template matching, in order to obtain more accurate matching results, bilinear interpolation needs to be performed on the template image and search region selected in Step 3 to obtain a higher resolution subpixel-level image: Let... Represents the midpoint of the image The pixel values, if the midpoint of the original image is known. The pixel values ​​at four adjacent integer points, which define a rectangle, can be used for bilinear subpixel interpolation to obtain the pixel value of any non-integer point in the matrix, as illustrated in the diagram below. Figure 4 As shown.

[0020] First of all Performing two linear interpolations in the direction yields:

[0021] Then in Performing a linear interpolation along the direction yields: .

[0022] in These represent the x and y coordinates of a pixel in the bilinear interpolated image. This represents the coordinates of four pixels in the original image.

[0023] Step 5: Target Tracking: The optimal adaptive template matching algorithm is used to track the position of the target in the video, obtaining its displacement time history data. For example... Figure 3 As shown, measurement point 2 is explained in this example. The subpixel-level template image and search region of measurement point 2 obtained in step 4 are used to perform optimal adaptive template matching in each frame of the video. In this invention, two templates are used simultaneously for template matching: one is a fixed template, i.e., the template image after subpixel processing; the other is a dynamic template, i.e., the template at the optimal matching position in the video frame is selected as the dynamic template. In the first frame of the video, the dynamic template is initialized to the fixed template. At this time, the fixed template and the dynamic template are the same, and both are used simultaneously for template matching in the first frame of the video. The matching position and matching result are the same, and the image at the matching position is selected to update the dynamic image. In the second frame of the video, the fixed template and the dynamic template are used for matching simultaneously, and the matching results are compared. The matching position with the higher matching result is selected to update the dynamic template. Subsequent video frames are similarly processed, achieving optimal adaptive template matching, obtaining more accurate matching results, avoiding the impact of target loss in video frames on subsequent matching results, and thus obtaining the displacement time history data of measurement point 2. And draw its displacement time history diagram based on the displacement time history data, such as Figure 4 As shown. Among them This represents the total number of video frames. Indicates the second measuring point Frame displacement.

[0024] The matching criterion used in template matching is standard correlation coefficient matching, which can be expressed as:

[0025] in Click on the template image Pixel value at the location, For the search area point Pixel value at the location, and These are the width and height of the template image, respectively. Click on the template image in the search area The matching results at the specified location.

[0026] Step 6: Spectrum Analysis: Displacement time history data of measuring point 2 A Fast Fourier Transform (FFT) is performed, which converts a discrete time-domain signal into a frequency-domain signal. This allows the determination of the natural frequencies of the cable. And draw a spectrum diagram, such as Figure 5 As shown. Among them This represents the total number of video frames, i.e., the length of the displacement time history signal. Indicates the second measuring point The frequency corresponding to each displacement signal.

[0027] Step 7: Cable Force Identification: Substitute the obtained cable natural frequency into the frequency-cable force formula to calculate the cable force, which can be expressed as:

[0028] For a slender cable, neglecting the effects of cable stiffness and sag, its cable force can be expressed as:

[0029] in For cable tension, The mass per unit length of the cable, The length of the cable, For Lasso's First natural frequency, The vibration order is... This refers to the bending stiffness of the cable. In practical engineering, a suitable frequency-cable force formula should be selected based on the actual situation and accuracy requirements.

[0030] This invention employs an infrared-based image acquisition device to capture vibration videos of cables, solving the problems of high difficulty and poor accuracy in cable force testing under adverse weather conditions (such as rain, fog, and snow) and the difficulty in conducting cable force testing in low-light environments (such as nighttime or dark environments) using existing machine vision technology. It utilizes an image noise reduction algorithm to reduce image noise, followed by a bilinear interpolation algorithm to improve image resolution, and an optimal adaptive template matching algorithm for target tracking to obtain the cable's displacement-time history curve. Fast Fourier transform is then used to perform spectral analysis on the displacement-time history data to obtain the cable's natural frequency, and the cable force is calculated using the force-frequency formula.

[0031] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for testing cable tension based on infrared light and optimal adaptive template matching, characterized in that, Includes the following steps: Step 1: Collect cable vibration video: Install infrared light targets at the measuring points on the cable, and then use a camera equipped with an infrared filter to collect the vibration video of the cable. Step 2: Image preprocessing: For the acquired cable vibration video, the video frames are first converted into grayscale images using an image preprocessing algorithm, and then the images are denoised. Step 3: Select template image and search area: In the first frame of the video, select the template image and its search area respectively; Step 3 further includes the following: the video frame image after image preprocessing contains a cable and at least one infrared target light at the measurement point location; the measurement point number is... In the first frame of the video, the infrared light target on the measuring point is selected as the template image, and the vibration range of the measuring point on the cable is analyzed. Then, the search area of ​​the measuring point is selected as the search area for each frame in the subsequent template matching. Step 4: Subpixel processing: Perform bilinear interpolation on the template image and the search region to obtain a higher resolution subpixel-level image; Step 5: Target tracking: Use the optimal adaptive template matching algorithm to track the position of the target in the video and obtain its displacement time history data; Step 5 further includes the following: two templates are used for template matching. One is a fixed template, which is the template image after subpixel processing; the other is a dynamic template, which is the template selected at the best matching position in the video frame. In the first frame of the video, the dynamic template is initialized to the fixed template. At this time, the fixed template and the dynamic template are the same. Both are used to perform template matching on the first frame of the video. The matching position and the matching result are the same. The image at the matching position is selected to update the dynamic image. In the second frame of the video, the fixed template and the dynamic template are used to perform matching simultaneously. The matching results of the two are compared. The matching position with the higher matching result is selected to update the dynamic template. Subsequent video frames are similarly used to achieve optimal adaptive template matching, obtain more accurate matching results, obtain displacement time history data, and draw a displacement time history diagram. Step 6: Spectrum analysis: Perform a fast Fourier transform on the displacement time history data to obtain the natural frequencies of the cable. Step 6 further includes the following: performing a fast Fourier transform on the displacement time history data to obtain the natural frequencies of the cable. Step 7: Cable Force Identification: Substitute the obtained cable natural frequency into the frequency-cable force formula to calculate the cable force; Step 7 further includes the following: substituting the obtained cable natural frequency into the frequency-cable force formula to calculate the cable force, expressed as: ； For a slender cable, neglecting the effects of cable stiffness and sag, its cable force can be expressed as: ； in For cable tension, The mass per unit length of the cable, The length of the cable, For Lasso's First natural frequency, The vibration order is... This refers to the bending stiffness of the cable.

2. The cable tension testing method based on infrared light and optimal adaptive template matching according to claim 1, characterized in that: Step 1 further includes the following: installing a certain number of infrared light targets at the measuring points on the cable, then installing an infrared filter on the camera, then turning on the infrared light switch, and using the camera to collect video of the cable's vibration.

3. The cable tension testing method based on infrared light and optimal adaptive template matching according to claim 1, characterized in that: Step 2 further includes the following: importing the acquired cable vibration video into a computer device, using an image preprocessing algorithm to convert the RGB three-channel video frame image into a single-channel grayscale image to speed up subsequent program calculations, and then performing denoising processing on the image to reduce image noise; the specific implementation of converting to grayscale is as follows: ； in This indicates that the x-coordinate of the original color image is... The vertical axis is The pixels; Represents pixels in a grayscale image grayscale value, Represents pixels in the original color image The pixel value of the red channel, Represents pixels in the original color image The pixel value of the green channel, Represents pixels in the original color image The pixel value of the blue channel.

4. The cable tension testing method based on infrared light and optimal adaptive template matching according to claim 1, characterized in that: Step 4 further includes the following: Let... Represents the midpoint of the image The pixel values, if the midpoint of the original image is known. The pixel values ​​at four adjacent integer points, which define a rectangle, can be used to perform bilinear subpixel interpolation to obtain the pixel value of any non-integer point in the matrix; First of all Performing two linear interpolations in the direction yields: ； ； Then in Performing a linear interpolation along the direction yields: ； in These represent the x and y coordinates of a pixel in the bilinear interpolated image. This represents the coordinates of four pixels in the original image.

5. A cable tension testing device based on infrared light and optimal adaptive template matching, performing the method described in any one of claims 1-4, characterized in that: Cable vibration video acquisition module: Infrared light targets are installed at the measuring points on the cable, and then a camera equipped with an infrared filter is used to acquire the vibration video of the cable. Image preprocessing module: For the acquired cable vibration video, the image preprocessing algorithm first converts the video frames into grayscale images, and then performs noise reduction processing on the images. The template image and search area selection module allows you to select the template image and its search area within the first frame of the video. Subpixel processing module: Performs bilinear interpolation on the template image and the search area to obtain a higher resolution subpixel-level image; Target tracking module: Uses the optimal adaptive template matching algorithm to track the position of the target in the video and obtain its displacement time history data; Spectrum analysis module: Performs fast Fourier transform on displacement time history data to obtain the natural frequencies of the cable. Cable force identification module: Substitutes the obtained cable natural frequency into the frequency-cable force formula to calculate the cable force.

6. A bridge monitoring device, characterized in that, It includes a processor and a memory; the memory stores computer-readable instructions, and the processor is configured to execute the computer-readable instructions, wherein the computer-readable instructions, when executed, perform the method according to any one of claims 1 to 4.

Images