A method for extracting structural vibration signals under adverse environmental interference

Abstract

The present invention relates to the technical field of visual measurement and vibration signal processing, and specifically to a method for extracting structural vibration signals under adverse environmental interference. The method comprises: collecting structural vibration images through a camera, preprocessing the images and constructing a space-time matrix, and using singular value decomposition to achieve noise reduction. The space-time matrix is ​​compressed into a one-dimensional vibration time series through displacement field estimation and spatial averaging. According to the delayed embedding method, the one-dimensional vibration time series is reconstructed into a two-dimensional phase space matrix, and singular value decomposition is applied again to achieve effective separation of the real vibration signal and the camera motion interference signal. By introducing singular value decomposition and embedded reconstruction technology, accurate and efficient structural vibration signal extraction is achieved, meeting the requirements for vibration data accuracy and real-time performance in engineering applications.

Description

Technical Field

[0001] The present invention relates to the technical field of visual measurement and vibration signal processing, and in particular to a method for extracting structural vibration signals under adverse environmental interference. Background Art

[0002] Structural vibration response is one of the key indicators in the safety assessment and health monitoring of engineering structures. It can intuitively reflect the dynamic characteristics and stability of the structure under external excitation. Current structural vibration monitoring technologies can be roughly divided into two categories: contact and non-contact. Contact vibration measurement includes accelerometers, laser vibrometers, etc. Although the measurement accuracy is high, the measurement points are limited and the sensor installation process may cause disturbances to the structure; laser vibrometers measure surface vibration velocity or displacement in a non-contact manner and have good measurement sensitivity, but are usually only suitable for small area or single point measurements, and have high requirements for environmental stability. Non-contact visual vibration measurement, including methods for extracting structural surface motion trajectories based on video image sequences, can achieve wide-area, multi-point, non-invasive monitoring, but its measurement accuracy is limited by image quality, acquisition equipment stability, and the robustness of subsequent processing algorithms.

[0003] Although existing visual measurement technology has achieved non-contact, multi-point vibration data acquisition to a certain extent, it still has the following major shortcomings: during the image acquisition process, the camera is easily interfered with by environmental factors such as wind and ground vibration, resulting in the superposition of camera motion signals in the image sequence, affecting the extraction of true vibration signals; traditional methods often assume that the camera is stable or rely solely on post-processing filtering, making it difficult to effectively distinguish between structural vibration and camera motion in complex environments; data processing algorithms are sensitive to noise, especially in low signal-to-noise ratio scenarios, where vibration frequency and modal information are easily confused or masked, affecting the accuracy of monitoring results.

[0004] The key to achieving high-precision structural vibration signal extraction lies in improving image signal processing capabilities, optimizing signal separation algorithms, and incorporating theories such as embedding theorems for data reconstruction to enhance the separability of vibration signal features from interference signals. Furthermore, developing a simple and efficient data processing workflow and visualization interface is crucial to ensuring the feasibility and scalability of this method for practical engineering applications. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention aims to provide a method for extracting structural vibration signals under adverse environmental interference, overcoming the shortcomings of traditional visual vibration measurement methods, such as insufficient processing of camera motion interference, low vibration signal extraction accuracy, and poor environmental adaptability. Compared with existing visual vibration monitoring methods, this invention overcomes the limitations of traditional image processing methods in noise suppression and signal separation by introducing singular value decomposition and embedded reconstruction technology. This method can achieve high-precision structural vibration signal extraction, especially in complex environments and low signal-to-noise ratio conditions, meeting the requirements for vibration data accuracy and real-time performance in engineering applications.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for extracting structural vibration signals under adverse environmental interference, the process is as follows:

[0008] Step 1: Configure the image acquisition system: Set the camera's pixel resolution to 1920 × 1080 pixels and the frame rate to 50 frames per second to ensure that the smallest dynamic motion details on the surface of the structure can be captured while also meeting the requirements of real-time sampling. The camera is fixed in front of the free end of the cantilever beam to ensure that the detection area covers the entire vibration amplitude range. A high-power fan is used to apply external excitation to the camera to cause it to shake.

[0009] Step 2: Image acquisition: Give the cantilever beam an initial excitation to allow it to vibrate freely, and use a camera to continuously capture vibration images for approximately 40 seconds to ensure a sufficient time window for modal identification.

[0010] Step 3: Image preprocessing: grayscale and normalize the vibration image to enhance signal characteristics and reduce the impact of background illumination changes;

[0011] Step 4: Spatiotemporal noise reduction: The vibration image sequence is constructed into a three-dimensional matrix. The singular value decomposition (SVD) method is applied to screen out singular modes that retain more than 90% of the energy, effectively removing random noise and improving image quality.

[0012] Step 5: Displacement field estimation and spatial averaging: Calculate the local block displacement field between adjacent frames using the cross-correlation method, perform spatial averaging in the region of interest, and compress it into a one-dimensional vibration time series.

[0013] Step 6: Phase space reconstruction: Based on the delayed embedded method, the one-dimensional vibration time series is reconstructed into a two-dimensional phase space matrix to fully reveal the dynamic characteristics of the signal;

[0014] Step 7: Vibration signal extraction and processing: Apply SVD again to retain 2-3 main singular modes (usually the main vibration modes of the structure) to effectively separate the true vibration signal of the structure from the camera motion interference signal;

[0015] Step 8: Signal reconstruction and analysis: The extracted main vibration mode is inversely transformed into a time domain signal. A benchmark modal test is performed on the cantilever beam using a laser vibrometer. The time domain vibration signal is compared and verified with the laser vibrometer results.

[0016] Step 9. Data Output and Visualization: Visualize the dynamic response of the cantilever beam through vibration modal animation. Export the processed vibration data and generate an analysis report that includes modal frequencies, vibration shape characteristics, and signal quality assessment.

[0017] Furthermore, in the data preprocessing step, the image is grayscaled and normalized. The specific method is as follows:

[0018] Using the formula Calculate the pixel value of the grayscale image, where Represents the color intensity of the red, green and blue channels respectively. ;

[0019] Using the formula Calculate the normalized pixel value, where is the pixel value of the grayscale image, min(Gray) and max(Gray) are the minimum and maximum values ​​of the grayscale image pixel value, is the target normalized range.

[0020] Furthermore, in the spatiotemporal noise reduction processing step, a spatiotemporal matrix is ​​constructed. The specific process includes:

[0021] The image sequence collected by the camera contains K frames, and the size of each frame is H×W, where H is the height and W is the width. For each frame, use the formula Flatten the two-dimensional image into a one-dimensional vector, where represents the pixel intensity vector of the kth frame of the image, Represents the k-th frame image.

[0022] The vectors of all frames are stacked column by column to form a The space-time matrix M, , where rows represent spatial dimensions (each pixel location) and columns represent temporal sequences (different frames).

[0023] Furthermore, the space-time matrix M is subjected to singular value decomposition, which specifically includes the following process:

[0024] Using the formula Transform the matrix, where represents the left singular vector (time basis), represents a diagonal matrix of singular values, represents the right singular vector (modal direction);

[0025] Take the first two singular values and the corresponding vector ;

[0026] Using the formula Reconstructing low-rank matrix ;

[0027] Based on the cumulative energy ratio, the main modal components corresponding to the first r singular values ​​with energy contribution rates greater than 90% are retained, and environmental noise and low-amplitude disturbances are filtered out. Calculate the cumulative energy contribution rate of the first r singular values, where p is the diagonal matrix in SVD The number of diagonal elements of , that is, the number of singular values;

[0028] Find the smallest r such that ,in , is the preset energy threshold.

[0029] Furthermore, the specific method of displacement field estimation is:

[0030] From image grayscale Select Reference subwindow , using the formula Select by coordinates Centered at The reference subwindow of

[0031] Recommended size is Pixels (empirical value, depends on image resolution and texture size).

[0032] Furthermore, the specific method of spatial averaging is:

[0033] Use the normalized cross-correlation calculation formula Calculate the current window With target frame The similarity of the candidate regions in , where To search for the offset vector, is the target frame The mean of is the target frame The mean of the structure vibration is manually or automatically selected as the spatial average domain, and finally a one-dimensional vibration time series is obtained. .

[0034] Furthermore, phase space reconstruction specifically includes the following processes:

[0035] According to the delay embedding method, the delay time , embedding dimension d = 10 ~ 10000, construct the vibration time series into a two-dimensional phase space matrix. Since the camera frame rate is 50 frames / second and the sampling rate is low, the delay time is selected , revealing the dynamic characteristics of the signal to the greatest extent; embedding dimension d = 10 ~ 10000 to construct a sufficiently large dimensional matrix for SVD to separate mixed signals.

[0036] For one-dimensional vibration time series , determine its delay time and embedding dimension d, forming The two-dimensional matrix is ​​as follows:

[0037] .

[0038] Furthermore, the vibration signal extraction and processing specifically includes the following steps:

[0039] The reconstructed phase space matrix was subjected to SVD again. Based on experience and experimental results, the first two singular values ​​often contribute more than 90% of the total energy and correspond to the singular modes of the main structural vibration signal. Therefore, only the first two singular modes were retained, corresponding to the cantilever beam's true vibration signal and the camera's subtle motion interference signal, respectively, to achieve effective signal separation.

[0040] Furthermore, the signal is reconstructed, and the specific method is as follows:

[0041] According to the first and second order modal signals of the cantilever beam separated by SVD, the time domain vibration signal is deduced. The time domain vibration signal of each mode corresponds to the amplitude change of the mode in time. The time domain vibration signal of the i-th mode can be expressed as a right singular matrix The i-th row of ,in , indicating a point in time.

[0042] Furthermore, data output and visualization specifically include the following processes:

[0043] Vibration modal animations intuitively display the dynamic response of cantilever beams, facilitating subsequent dynamic modeling and simulation analysis. Processed vibration data is exported and an analysis report automatically generated containing modal frequencies, vibration shape characteristics, and signal quality assessments, facilitating engineering applications and decision support.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] 1. This invention combines non-contact visual measurement with singular value decomposition and extraction technology to effectively achieve high-precision separation of the structure's true vibration signal and camera interference signals in complex environments. This can significantly improve the reliability and accuracy of vibration signal extraction, especially at construction sites with severe wind disturbances, ground vibrations, or equipment interference.

[0046] 2. By introducing a two-stage singular value decomposition process (spatiotemporal noise reduction + phase space separation), this invention can preserve the main vibration modes of the structure in a high-noise environment and suppress random noise and spurious signals, ensuring the complete extraction of vibration characteristics. The system also supports real-time visualization and analysis of vibration data, generating detailed modal frequency and dynamic response diagrams, and supports data export and automated generation of standardized analysis reports, facilitating subsequent engineering evaluation and decision support.

[0047] 3. No complex calculations are involved in the signal conversion process, which optimizes the data processing process, takes into account both algorithm efficiency and accuracy, and meets the real-time and large-scale data processing requirements in engineering applications;

[0048] 4. The method has a wide range of applications, covering vibration monitoring tasks for different types of structures (such as buildings, bridges, tunnels, and mechanical equipment) and under various excitation conditions (such as wind vibration, traffic loads, and earthquakes). BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0050] Figure 1 This is a flow chart of a method for extracting structural vibration signals under adverse environmental interference;

[0051] Figure 2 and Figure 3 2 is a data comparison chart of the test results in the embodiments of the present invention. DETAILED DESCRIPTION

[0052] Example

[0053] See Figure 1 As shown, a method for extracting structural vibration signals under adverse environmental interference in this embodiment has the following process:

[0054] Step 1: Configuration of image acquisition system.

[0055] The camera's pixel resolution was set to 1920×1080 pixels, with a frame rate of 50 frames per second, ensuring the capture of minute dynamic motion details on the structure's surface while also meeting real-time sampling requirements. The camera was fixed in front of the free end of the cantilever beam, ensuring the detection area covered the entire vibration amplitude range. A high-power fan was used to apply external excitation to the camera, causing it to shake.

[0056] Step 2: Image acquisition.

[0057] An initial excitation is given to the cantilever beam to allow it to vibrate freely, and vibration images are continuously collected for about 40 seconds using a camera to ensure a sufficient time window for modal identification.

[0058] Step 3: Image preprocessing (eliminating interference caused by background lighting changes and enhancing signal characteristics).

[0059] Grayscale processing: The weighted average method is used for grayscale processing. The specific formula is as follows:

[0060] ;

[0061] in, Represents the color intensity of the red, green and blue channels respectively. .

[0062] Normalization correction: Map the grayscale image pixel values ​​to the specified interval. The specific formula is as follows:

[0063] ;

[0064] in, is the pixel value of the grayscale image, min(Gray) and max(Gray) are the minimum and maximum values ​​of the grayscale image pixel value, is the target normalization range, in this embodiment .

[0065] Step 4: Spatiotemporal noise reduction processing.

[0066] Construct the space-time matrix:

[0067] The vibration image sequence is constructed into a spatiotemporal matrix according to the following method:

[0068] The image sequence captured by the camera contains K frames, each of which has a size of H×W, where H is the height and W is the width. For each frame, the two-dimensional image is flattened into a one-dimensional vector using the following formula:

[0069] ;

[0070] in, represents the pixel intensity vector of the kth frame of the image, Represents the k-th frame image.

[0071] The vectors of all frames are stacked column by column to form a The space-time matrix is ​​as follows:

[0072] ;

[0073] Here, rows represent spatial dimensions (each pixel location) and columns represent temporal sequences (different frames).

[0074] Singular Value Decomposition (SVD):

[0075] Perform singular value decomposition on the above space-time matrix M. The specific formula is as follows:

[0076] ;

[0077] : left singular vector (time basis);

[0078] : singular value diagonal matrix;

[0079] : right singular vector (modal direction);

[0080] Take the first two singular values and the corresponding vector ;

[0081] Reconstructing low-rank matrix , the formula is as follows:

[0082] ;

[0083] Based on the cumulative energy ratio, the main modal components corresponding to the first r singular values ​​with an energy contribution rate greater than 90% are retained, and environmental noise and low-amplitude disturbances are filtered out. The cumulative energy contribution rate of the first r singular values ​​is calculated as follows:

[0084] ;

[0085] Where p is the diagonal matrix in SVD The number of diagonal elements, that is, the number of singular values.

[0086] The goal is to find the smallest r such that ,in , is the preset energy threshold. Based on prior knowledge, in this embodiment, r=2.

[0087] Step 5: Displacement field estimation and spatial averaging.

[0088] Displacement field estimation: from image grayscale Select Reference subwindow , the following is a selection with coordinates Centered at The formula for the reference subwindow is:

[0089] ;

[0090] Recommended size is Pixels (empirical value, depends on image resolution and texture scale); can overlap sliding samples by 50% to form a dense displacement field.

[0091] Spatial averaging: Calculates the current window using normalized cross-correlation With target frame The similarity of the candidate regions in , the formula is as follows:

[0092] ;

[0093] in, To search for the offset vector, is the target frame The mean of is the target frame The mean of .

[0094] Manually or automatically select the significant structural vibration area as the spatial average domain, and finally obtain the one-dimensional vibration time series .

[0095] Step 6: Phase space reconstruction.

[0096] According to the delay embedding method, the delay time , embedding dimension d = 10~10000, constructing the vibration time series into a two-dimensional phase space matrix, suitable for subsequent extraction and analysis. In this embodiment, the frame rate of the camera is 50 frames / second, and the sampling rate is low, so the delay time is selected , revealing the dynamic characteristics of the signal to the greatest extent; embedding dimension d = 10 ~ 10000 to construct a matrix of sufficiently large dimensions for SVD to separate mixed signals. The details are as follows:

[0097] For one-dimensional vibration time series , determine its delay time and embedding dimension ,form A two-dimensional matrix:

[0098] ;

[0099] Step 7: Vibration signal extraction and processing.

[0100] The phase space matrix obtained in step 6 is subjected to SVD again. Based on experience and experimental results, the first two singular values ​​often contribute more than 90% of the total energy and correspond to the singular modes of the main structural vibration signal. Therefore, only the first two singular modes are retained, corresponding to the cantilever beam's true vibration signal and the camera's micro-motion interference signal, respectively, to effectively extract the vibration signal.

[0101] Step 8: Signal reconstruction and analysis.

[0102] The time domain vibration signal is inferred from the first and second order modal signals of the cantilever beam separated by SVD. The time domain vibration signal of each mode corresponds to the amplitude change of the mode over time. The specific calculation method is as follows:

[0103] The time domain vibration signal of the i-th mode is a right singular matrix The i-th row of:

[0104] ;

[0105] in , indicating a point in time.

[0106] To verify the accuracy of the extracted signal, a laser vibrometer was used to perform a benchmark modal test on the cantilever beam. The time domain vibration signal was compared with the results of the laser vibrometer, as shown in the figure. Figure 2 is the structural vibration signal measured by the camera under wind load, Figure 3 is the structural vibration signal measured by the camera under the action of seismic waves. The normalized root mean square error (NRMSE) between the traditional method and the method of the present invention is shown in Figure 2 and Figure 3 Legend for .

[0107] Step 9: Data output and visualization.

[0108] The extracted time-domain vibration signal of the structure can be input into Matlab or Python software, and the dynamic response of the cantilever beam can be intuitively displayed through vibration modal animation, which is convenient for subsequent dynamic modeling and simulation analysis. At the same time, the processed vibration data can be exported to generate an analysis report containing modal frequency, vibration shape characteristics and signal quality assessment, which is convenient for engineering application and decision support.

[0109] Through the detailed introduction and comparative analysis of this example, the vibration signal extraction method based on SVD spatiotemporal noise reduction and extraction demonstrates significant advantages over traditional visual measurement methods in multiple aspects, particularly in vibration signal extraction accuracy in complex environments, non-contact operation, and data processing efficiency. This method provides a more reliable, economical, and efficient technical solution for monitoring dynamic characteristics in practical engineering applications.

[0110] The above formulas are all dimensionless and numerically calculated, and the preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0111] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0112] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0113] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0114] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0116] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0117] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for extracting structural vibration signals under adverse environmental interference, characterized in that: include: Step 1: Image acquisition system configuration: Fix the camera in front of the structure to be measured and use a fan to apply external excitation to the camera to make it shake; Step 2: Image acquisition: Apply an initial excitation to the structure to be tested, allowing it to vibrate freely, and use a camera to continuously capture vibration images of the structure to be tested; Step 3: Image preprocessing: grayscale and normalize the vibration image of the structure to be measured; Step 4: Spatiotemporal noise reduction: The vibration image of the structure to be measured is constructed into a three-dimensional matrix, and the singular value decomposition method is applied to perform noise reduction. The specific process of constructing the three-dimensional matrix includes: The image sequence collected by the camera contains K frames of images, each frame has a size of H×W, where H is the height and W is the width; for each frame of image, use the formula , flatten the two-dimensional image into a one-dimensional vector, where represents the pixel intensity vector of the kth frame of the image, Represents the k-th frame image; the vectors of all frames are stacked in columns to form a The space-time matrix M, , where rows represent spatial dimensions and columns represent time series; Step 5: Displacement field estimation and spatial averaging: Calculate the local block displacement field between adjacent frames in the constructed three-dimensional matrix using the cross-correlation method and perform spatial averaging to obtain the one-dimensional vibration time series of the structure under test; Step 6: Phase space reconstruction: Based on the delayed embedded method, the one-dimensional vibration time series is reconstructed into a two-dimensional phase space matrix, which specifically includes the following steps: According to the delay embedding method, the delay time , embedding dimension d = 10 ~ 10000, constructing the one-dimensional vibration time series into a two-dimensional phase space matrix; for the one-dimensional vibration time series , determine its delay time and embedding dimension ,form The following two-dimensional matrix: ； Step 7: Extraction and processing: Apply singular value decomposition to the reconstructed phase space matrix again, retain the main singular modes, and separate the true vibration signal of the structure to be measured from the camera motion interference signal; Step 8: Signal reconstruction and analysis: The main singular modes obtained by singular value decomposition are inversely transformed into time domain signals. A laser vibrometer is used to perform a benchmark modal test on the structure under test. The measurement results are compared and analyzed with the vibration signals extracted by this method. Step 9. Data output and visualization: Vibration modal animation is used to visually display the dynamic response of the structure under test, and frequency and modal shape data files are exported.

2. The method for extracting structural vibration signals under adverse environmental interference according to claim 1, characterized in that: Perform singular value decomposition on the space-time matrix M, specifically The following processes are included: Using the formula , transform the matrix, where represents the left singular vector, represents a diagonal matrix of singular values, Represents the right singular vector; take the first two singular values and the corresponding vector ; Using the formula , reconstruct the low-rank matrix ; Taking the cumulative energy ratio as the criterion, retain the main modal components corresponding to the first r singular values ​​with energy contribution rates greater than 90%; use the formula , calculate the cumulative energy contribution rate of the first r singular values, where p is the diagonal matrix in the singular value decomposition The number of diagonal elements; find the smallest r such that ,in , is the preset energy threshold.

3. The method for extracting structural vibration signals under adverse environmental interference according to claim 1, characterized in that: The specific method of displacement field estimation and spatial averaging is: From pixel intensity vector Select Reference subwindow , using the formula , select the coordinate Centered at The reference subwindow of Use the normalized cross-correlation calculation formula , calculate the current window With target frame The similarity of the candidate regions in , where To search for the offset vector, is the target frame The mean of is the target frame The mean of Manually or automatically select the vibration-significant area of ​​the structure to be tested as the spatial average domain, and finally obtain a one-dimensional vibration time series .

4. The method for extracting structural vibration signals under adverse environmental interference according to claim 1, characterized in that: Extraction processing, specific The following processes are included: The reconstructed phase space matrix is ​​subjected to singular value decomposition again, and the first two singular modes are retained, which correspond to the real vibration signal of the structure to be measured and the slight motion interference signal of the camera, respectively.

5. The method for extracting structural vibration signals under adverse environmental interference according to claim 2, characterized in that: Signal reconstruction, the specific method is: According to the first-order and second-order modal signals of the structure to be tested separated by singular value decomposition, the time domain vibration signal is deduced; the time domain vibration signal of the i-th mode can be expressed as the right singular matrix The i-th row of ,in , indicating a point in time.

6. The method for extracting structural vibration signals under adverse environmental interference according to claim 1, characterized in that: Data output and visualization, including: The dynamic response of the structure to be tested is intuitively displayed through vibration modal animation; the processing system of this method supports the export of processed vibration data and can automatically generate an analysis report containing modal frequencies, vibration shape characteristics and signal quality evaluation.

Images