A mechanical vibration measurement device and method based on composite vibration fringes

Abstract

The present invention proposes a mechanical vibration measurement device and method based on composite vibration stripes, which relate to the technical field of mechanical vibration measurement. The vibration measurement device includes a composite vibration stripe target, a high-speed camera, and a computer; the composite vibration stripe target is attached to a mechanical structure to capture its vibration change information; the camera is used to capture images of the composite vibration stripe pattern on the attached target and send the captured images to a computer for analysis; the computer post-processes the vibration stripe image to obtain vibration information such as the frequency and amplitude of the mechanical structure. The method and device have the advantages of a wide detection field of view and high detection accuracy; they are suitable for use in mechanical structure vibration measurement tasks in various environments and have broad application prospects.

Description

Technical Field

[0001] The present invention relates to the field of visual vibration measurement technology, and in particular to a mechanical vibration measurement device and method based on composite vibration fringes. Background Art

[0002] In the existing mechanical system control environment, accurate measurement of mechanical vibration is very important to ensure stable operation and precise control of the equipment. Traditional vibration measurement methods mainly rely on hardware devices such as sensors and encoders. Although they can provide a certain degree of positioning accuracy, their equipment costs are high, and the layout and maintenance under complex working conditions are relatively cumbersome. On the other hand, vision-based vibration measurement methods are lower in cost, but they require the construction of complex spatial associations between the mechanical structure and the camera, and in certain application scenarios, they are easily affected by factors such as lighting and noise, and may have problems such as insufficient accuracy and poor real-time performance. Evaluating the vibration characteristics and control accuracy of mechanical structures under variable working conditions, and verifying their vibration positioning and response accuracy during actual operation are indispensable tasks.

[0003] Based on this, we proposed a mechanical vibration measurement device and method based on composite vibration stripes. By deploying designed composite vibration stripe targets at key parts of the mechanical structure under test and combining them with corresponding detection and decoding methods, we can achieve multi-dimensional acquisition and rapid identification of vibration information such as amplitude and frequency. Summary of the Invention

[0004] The purpose of the present invention is to provide a mechanical vibration measurement device and method based on composite vibration fringes, which has the characteristics of wide detection range and high detection accuracy.

[0005] To achieve the above object, the technical solution of the present invention is as follows:

[0006] The present invention proposes a mechanical vibration measurement method based on composite vibration fringes, comprising the following steps:

[0007] Step S1: attach the composite vibration fringe pattern to the mechanical structure to be tested, and establish a three-dimensional coordinate at the center point of the composite vibration fringe pattern, with the camera optical axis perpendicular to the camera imaging surface as the Z The horizontal direction perpendicular to the imaging optical axis is defined as the X-axis, and the vertical direction perpendicular to the imaging optical axis is defined as the Y-axis, which are used to sense the vibration signal of the mechanical structure; the high-speed camera is placed in a stable area, and the focal length and aperture size are set according to the sensor size and shooting distance so that the composite vibration fringe pattern can be captured by the high-speed camera;

[0008] Step S2: When the mechanical structure vibrates, the relative position of the composite vibration fringe pattern attached to the mechanical structure with respect to the imaging sensor of the high-speed camera also changes accordingly; a vibration fringe image with multi-dimensional vibration signals is acquired by the high-speed camera;

[0009] Step S3: transmitting the collected vibration fringe signals to a computer, which decodes the vibration fringe signals and extracts multi-vibration information of the mechanical structure; specifically, the following steps are included:

[0010] Step S31: For each frame of the vibration fringe image, the Pearson correlation coefficient between the grayscale value of the calculation area of the vibration fringe image and the grayscale value of the reference area in the X-axis and Y-axis directions is calculated, and a Pearson correlation coefficient curve is obtained; with the initial frame as the reference frame, the Pearson correlation coefficient curves of the reference frame and the vibration frame are cross-correlated in the X-axis and Y-axis directions at different vibration offset values, respectively, to obtain the vibration offset value index corresponding to the maximum cross-correlation value between the reference frame and the vibration frame in the X-axis and Y-axis directions, and the two-dimensional vibration information of the mechanical structure in the X-axis and Y-axis directions in the vibration fringe images of the reference frame and the vibration frame is calculated based on the obtained vibration offset value index;

[0011] Step S32: For each frame of the vibration fringe image, the fringe frequency is calculated using a frequency correction method; and based on a mathematical model of frequency and object distance, the single-dimensional vibration information of the mechanical structure in the Z-axis direction in the current frame of the vibration fringe image is calculated;

[0012] Step S33: according to the grayscale information of each frame of the vibration fringe image, a vibration fringe decoding strategy based on a Gabor filter is used to identify the in-plane rotation angle information of the mechanical structure;

[0013] Step S34: Calculate the object distance based on the obtained row and column fringe frequencies and the mathematical model of frequency and object distance, and obtain the corrected object distance through data fitting; derive a mathematical model between the composite vibration-sensing fringe pattern and the vibration of the mechanical structure to obtain the out-of-plane rotation angle information of the mechanical structure based on the corrected object distance.

[0014] Preferably, the composite vibration-sensitive stripe pattern is based on sinusoidal stripes, and by modulating the frequency and direction information of the sinusoidal stripes, a composite coded stripe pattern of the same frequency in different directions is formed, including a single sinusoidal stripe area in the middle and surrounding oblique stripe areas;

[0015] The middle single sinusoidal stripe region includes four square single sinusoidal stripe regions, the edges of the four single sinusoidal stripe regions are respectively connected to the edges of the middle square gray block and form a cross with the gray block, and the stripe directions of the four single sinusoidal stripe regions are respectively parallel to the edges of the gray block;

[0016] The surrounding oblique stripe area includes four square oblique stripe areas located in the blank areas around the middle single sinusoidal stripe area. The stripes of the four square oblique stripe areas have the same direction and are connected to the stripes of the single sinusoidal stripe area.

[0017] Preferably, the Pearson correlation coefficients between the grayscale values of the calculation area and the grayscale values of the reference area of the vibration fringe images in the X-axis and Y-axis directions are calculated respectively, as follows:

[0018] When calculating the Pearson correlation coefficient between the grayscale values of the calculation area and the grayscale values of the reference area of the vibration fringe image in the X-axis direction, the two single sinusoidal fringe areas with the stripes perpendicular to the X-axis and the square gray block in the middle are used as the reference areas, and the other areas except the reference area are used as the calculation areas;

[0019] Take the grayscale value of any row in the reference area as the reference row grayscale value and calculate the Pearson correlation coefficient:

[0020]

[0021] in Is the reference line r Gray value and calculation area i PCC correlation coefficient of row grayscale value; I r is the grayscale value of the reference row, I i The calculation area is i Row grayscale value; yes I r and I i covariance of yes I r The standard deviation of yes I i The standard deviation of Is the reference line r No. j Column grayscale value, Calculation area i Rank j Column grayscale value, n is the number of columns; and They are I r and I i The mean of

[0022] When calculating the Pearson correlation coefficient between the grayscale values of the calculation area and the reference area of the vibration fringe image in the Y-axis direction, the two single sinusoidal fringe areas with the stripes perpendicular to the Y-axis and the square gray block in the middle are used as the reference areas, and the other areas except the reference area are used as the calculation areas;

[0023] Take any column grayscale value in the reference area as the reference column grayscale value and calculate the Pearson correlation coefficient:

[0024]

[0025] in Is a reference column c Gray value and calculation area j PCC correlation coefficient of column gray value; I c is the grayscale value of the reference column, I j The calculation area is j Column grayscale value; yes I c and I j covariance of yes I c The standard deviation of yes I j The standard deviation of Is a reference column c No. i Row grayscale value, Calculation area j Liedi i Row grayscale value, m is the number of rows; and They are I c and I j The mean of .

[0026] Preferably, cross-correlation calculation is performed on the Pearson correlation coefficient curves of the reference frame and the vibration frame in the X-axis and Y-axis directions at different vibration offset values, and the vibration offset value index corresponding to the maximum cross-correlation value between the reference frame and the vibration frame in the X-axis and Y-axis directions is obtained. The two-dimensional vibration information of the mechanical structure in the X-axis and Y-axis directions in the vibration fringe images of the reference frame and the vibration frame is calculated based on the obtained vibration offset value index; specifically as follows:

[0027] After Fourier transforming the obtained Pearson correlation coefficient curves in the X-axis and Y-axis directions, the Fourier transformed correlation coefficient curves are upsampled by zero padding in the frequency domain, and then the time domain information is returned by inverse Fourier transform to increase the pixel resolution of the vibration measurement. The cross-correlation calculation is performed using the processed Pearson correlation coefficient curves in the X-axis and Y-axis directions of the reference frame and the vibration frame:

[0028]

[0029] in h 1[ s ] is the reference frame signal in the X-axis or Y-axis direction h 1 in data point s The Pearson correlation coefficient value at h 2[ s+k ] is the vibration frame signal corresponding to the X-axis or Y-axis direction h 2 in data point s + k The Pearson correlation coefficient value at S is the signal length; k is the vibration offset value, which can be positive, negative or zero; R 12 [ k ] is the reference frame and vibration frame signal in the X-axis or Y-axis direction h 1 and h 2 in vibration offset value k Cross-correlation value under ;

[0030] After obtaining the vibration offset value index k corresponding to the maximum cross-correlation value of the reference frame and the vibration frame in the X-axis and Y-axis directions, the peak correction method is used to correct the k , multiply the corrected offset in the corresponding direction by the scale factor to obtain the vibration magnitude in the X-axis and Y-axis directions;

[0031] The formula for calculating the scale factor in the X-axis or Y-axis direction is as follows:

[0032]

[0033] in, M is the scaling factor, L is the actual length of the composite vibration fringe pattern in the X-axis or Y-axis direction, a is the size of a single pixel in the imaging camera, r' It is the number of pixels on the camera sensor occupied by the side length of the vibration fringe image in the X-axis or Y-axis direction.

[0034] Preferably, for each frame of vibration fringe image, the fringe frequency is calculated using a frequency correction method; specifically:

[0035] The grayscale values of the vibration fringe image are subjected to Fourier transformation row by row and column by column to obtain the fringe frequency of each row and column in the vibration fringe image. The obtained fringe frequency of each row and column is corrected using the peak correction method to obtain multiple groups of fringe frequencies of the corrected vibration fringe image in the X-axis and Y-axis directions.

[0036] Preferably, the mathematical model based on frequency and object distance is used to calculate the single-dimensional vibration information of the mechanical structure in the Z-axis direction in the vibration fringe image of the current frame; specifically:

[0037] Construct a mathematical model of the fringe frequency and the distance between targets along the Z axis of the camera's optical axis:

[0038]

[0039] in, f I is the fringe frequency of the vibration fringe image on the sensor, D is the distance between the imaging camera and the composite vibration fringe pattern along the camera optical axis Z axis, F is the focal length of the imaging lens, a is the size of a single pixel in the imaging camera, p is the fringe period of the composite vibration-sensitive fringe pattern;

[0040] For the multiple groups of fringe frequencies of the corrected vibration fringe image in the X-axis and Y-axis directions, the median of the fringe frequencies of each column in the X-axis direction and the median of the fringe frequencies of each row in the Y-axis direction are averaged to obtain the frequency of the vibration fringe image on the sensor. f I , substitute into the mathematical model to calculate the object distance D , obtain the position information of the mechanical structure in the Z-axis direction.

[0041] Preferably, the step S33 is specifically as follows:

[0042] Convolve the vibration-fringe image:

[0043]

[0044] in R ( x , y ) is the Gabor filter at pixel position ( x , y ) output response at; I ( u , v ) is the vibration fringe image at position ( u , v ), The spatial frequency is f, direction is i Gabor filter kernel;

[0045] The two-dimensional calculation formula of the Gabor filter is as follows:

[0046]

[0047] in is a two-dimensional Gabor function, Represents the coordinates after rotation transformation:

[0048]

[0049] in, represents the Gabor filter wavelength, and ; represents the Gaussian standard deviation; Indicates the spatial aspect ratio; represents phase shift; represents an imaginary unit;

[0050] The Gabor filter wavelength Set to the fringe frequency in the vibration fringe image f I The inverse of the Gabor filter direction i The value range is set to 0-360°, and the step size is set to between 0.01° and 1°. According to the preset step size, all directions of the Gabor filter are traversed and the filter output response of each direction is calculated. The Gabor filter angle corresponding to the maximum output response is i , as the in-plane rotation angle of the mechanical structure in the Z-axis direction.

[0051] Preferably, the object distance is calculated based on the obtained frequency of each row and column fringe and the mathematical model of frequency and object distance, and the corrected object distance is obtained by data fitting; specifically,

[0052] In the X-axis direction, based on the obtained frequency of each fringe column, the corresponding object distance is calculated according to the mathematical model of frequency and object distance. D x1 , D x2 ,… D xc , and obtain the corrected object distance by least square fitting D' x1 , D' x2 ,… D' xc ;

[0053] In the Y-axis direction, based on the obtained frequency of each line of stripes, the corresponding object distance is calculated according to the mathematical model of frequency and object distance. D y1 , D y2 ,… D yr , and obtain the corrected object distance by least square fitting D' y1 , D' y2 ,… D' yr .

[0054] Preferably, the mathematical model derived between the composite vibration-sensitive fringe pattern and the mechanical structure vibration is used to obtain the out-of-plane rotation angle information of the mechanical structure according to the corrected object distance; specifically:

[0055] Out-of-plane rotation angle in the X-axis direction α Expressed as:

[0056]

[0057] in, D' x1 is the distance between the first row of fringe in the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction, D' xc is the distance between the last row of fringe in the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction; D is the distance between the imaging camera and the composite vibration fringe pattern along the camera optical axis Z-axis obtained in step S32; r x is the number of camera sensor pixels occupied by the composite vibration fringe pattern in the X-axis direction;

[0058] Out-of-plane rotation angle in the Y-axis direction β Expressed as:

[0059]

[0060] in, D' y1 is the distance between the first row of fringe in the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction, D' yr is the distance between the last row of the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction; r y is the number of camera sensor pixels occupied by the composite vibration fringe pattern in the Y-axis direction.

[0061] The present invention also provides a mechanical vibration measurement device based on composite vibration fringes, the device being implemented using any of the above-mentioned mechanical vibration measurement methods based on composite vibration fringes, comprising a composite vibration fringes target, a high-speed camera, and a computer;

[0062] The composite vibration fringe target is provided with a composite vibration fringe pattern, and the composite vibration fringe target is attached to the mechanical structure to be measured to obtain vibration sensing information;

[0063] The high-speed camera is responsible for collecting images of the composite vibration-sensitive fringe pattern and transmitting the images to a computer;

[0064] The computer extracts vibration perception information of the target image and performs analysis and processing to obtain multi-dimensional vibration information of the mechanical structure.

[0065] Compared with the prior art, the present invention has the following beneficial effects:

[0066] A mechanical vibration measurement device and method using composite vibration stripes is developed. By attaching a composite vibration stripe pattern to a mechanical structure, capturing motion images of the object to be measured using a high-speed camera, and finally analyzing and processing the vibration stripe images using a computer, multi-dimensional vibration information of the mechanical structure is obtained. Compared with existing algorithms, this calculation method has the advantages of a wide detection range and high measurement accuracy, and has extremely high practical value and application potential. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 This is a schematic structural diagram of the mechanical vibration measuring device of the composite vibration-sensing fringes of the present invention.

[0068] In the picture:

[0069] 1- High-speed camera; 2- Computer; 3- Composite vibration-sensitive fringe target; 4- Mechanical structure to be measured. DETAILED DESCRIPTION

[0070] The following is combined with Figure 1 , the technical solution of the present invention is described in detail.

[0071] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present application belongs.

[0072] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0073] The present invention proposes a mechanical vibration measurement method based on composite vibration fringes. Figure 1 , including the following steps:

[0074] Step S1: attach the composite vibration fringe pattern to the mechanical structure to be tested 4, and establish a three-dimensional coordinate at the center point of the composite vibration fringe pattern, with the camera optical axis perpendicular to the camera imaging surface as the Z Axis, the horizontal direction perpendicular to the imaging optical axis is defined as the X-axis, and the vertical direction perpendicular to the imaging optical axis is defined as the Y-axis, which are used to sense the vibration signal of the mechanical structure; the high-speed camera 1 is placed in a stable area, and the focal length and aperture size are set according to the sensor size and shooting distance so that the composite vibration fringe pattern can be captured by the high-speed camera 1;

[0075] Step S2: When the mechanical structure vibrates, the relative position of the composite vibration fringe pattern attached to the mechanical structure to the imaging sensor of the high-speed camera 1 also changes accordingly; the high-speed camera 1 acquires a vibration fringe image with a multi-dimensional vibration signal;

[0076] Step S3: transmitting the collected vibration fringe signals to the computer 2, which decodes the vibration fringe signals and extracts multi-dimensional vibration information of the mechanical structure; specifically, the following steps are included:

[0077] Step S31: For each frame of the vibration fringe image, the Pearson correlation coefficient between the grayscale value of the calculation area of the vibration fringe image and the grayscale value of the reference area in the X-axis and Y-axis directions is calculated, and a Pearson correlation coefficient curve is obtained; with the initial frame as the reference frame, the Pearson correlation coefficient curves of the reference frame and the vibration frame are cross-correlated in the X-axis and Y-axis directions at different vibration offset values, respectively, to obtain the vibration offset value index corresponding to the maximum cross-correlation value between the reference frame and the vibration frame in the X-axis and Y-axis directions, and the two-dimensional vibration information of the mechanical structure in the X-axis and Y-axis directions in the vibration fringe images of the reference frame and the vibration frame is calculated based on the obtained vibration offset value index;

[0078] Step S32: For each frame of the vibration fringe image, the fringe frequency is calculated using a frequency correction method; and based on a mathematical model of frequency and object distance, the single-dimensional vibration information of the mechanical structure in the Z-axis direction in the current frame of the vibration fringe image is calculated;

[0079] Step S33: according to the grayscale information of each frame of the vibration fringe image, a vibration fringe decoding strategy based on a Gabor filter is used to identify the in-plane rotation angle information of the mechanical structure;

[0080] Step S34: Calculate the object distance based on the obtained row and column fringe frequencies and the mathematical model of frequency and object distance, and obtain the corrected object distance through data fitting; derive a mathematical model between the composite vibration-sensing fringe pattern and the vibration of the mechanical structure to obtain the out-of-plane rotation angle information of the mechanical structure based on the corrected object distance.

[0081] In this embodiment, the composite vibration-sensitive stripe pattern is based on sinusoidal stripes. By modulating the frequency and direction information of the sinusoidal stripes, a composite coded stripe pattern with the same frequency in different directions is formed, including a single sinusoidal stripe area in the middle and surrounding oblique stripe areas.

[0082] The middle single sinusoidal stripe region includes four square single sinusoidal stripe regions, the edges of the four single sinusoidal stripe regions are respectively connected to the edges of the middle square gray block and form a cross with the gray block, and the stripe directions of the four single sinusoidal stripe regions are respectively parallel to the edges of the gray block;

[0083] The surrounding oblique stripe area includes four square oblique stripe areas located in the blank areas around the middle single sinusoidal stripe area. The stripes of the four square oblique stripe areas have the same direction and are connected to the stripes of the single sinusoidal stripe area.

[0084] In this embodiment, the Pearson correlation coefficients between the grayscale values of the calculation area and the grayscale values of the reference area of the vibration fringe images in the X-axis and Y-axis directions are calculated respectively, as follows:

[0085] When calculating the Pearson correlation coefficient between the grayscale values of the calculation area and the grayscale values of the reference area of the vibration fringe image in the X-axis direction, the two single sinusoidal fringe areas with the stripes perpendicular to the X-axis and the square gray block in the middle are used as the reference areas, and the other areas except the reference area are used as the calculation areas;

[0086] Take the grayscale value of any row in the reference area as the reference row grayscale value and calculate the Pearson correlation coefficient:

[0087]

[0088] in Is the reference line r Gray value and calculation area i PCC correlation coefficient of row grayscale value; I r is the grayscale value of the reference row, I i The calculation area isi Row grayscale value; yes I r and I i covariance of yes I r The standard deviation of yes I i The standard deviation of Is the reference line r No. j Column grayscale value, Calculation area i Rank j Column grayscale value, n is the number of columns; and They are I r and I i The mean of

[0089] When calculating the Pearson correlation coefficient between the grayscale values of the calculation area and the reference area of the vibration fringe image in the Y-axis direction, the two single sinusoidal fringe areas with the stripes perpendicular to the Y-axis and the square gray block in the middle are used as the reference areas, and the other areas except the reference area are used as the calculation areas;

[0090] Take any column grayscale value in the reference area as the reference column grayscale value and calculate the Pearson correlation coefficient:

[0091]

[0092] in Is a reference column c Gray value and calculation area j PCC correlation coefficient of column gray value; I c is the grayscale value of the reference column, I j The calculation area is j Column grayscale value; yes I c and I j covariance of yes I c The standard deviation of yes I j The standard deviation of Is a reference column c No. i Row grayscale value, Calculation area j Liedi i Row grayscale value, m is the number of rows; and They are I c and I j The mean of .

[0093] In this embodiment, cross-correlation calculations are performed on the Pearson correlation coefficient curves of the reference frame and the vibration frame in the X-axis and Y-axis directions at different vibration offset values, and vibration offset value indexes corresponding to the maximum cross-correlation values of the reference frame and the vibration frame in the X-axis and Y-axis directions are obtained. Based on the obtained vibration offset value indexes, two-dimensional vibration information of the mechanical structure in the X-axis and Y-axis directions in the vibration fringe images of the reference frame and the vibration frame is calculated; specifically, as follows:

[0094] After Fourier transforming the obtained Pearson correlation coefficient curves in the X-axis and Y-axis directions, the Fourier transformed correlation coefficient curves are upsampled by zero padding in the frequency domain, and then the time domain information is returned by inverse Fourier transform to increase the pixel resolution of the vibration measurement. The cross-correlation calculation is performed using the processed Pearson correlation coefficient curves in the X-axis and Y-axis directions of the reference frame and the vibration frame:

[0095]

[0096] in h 1[ s ] is the reference frame signal in the X-axis or Y-axis direction h 1 in data point s The Pearson correlation coefficient value at h 2[ s+k ] is the vibration frame signal corresponding to the X-axis or Y-axis direction h 2 in data point s + k The Pearson correlation coefficient value at S is the signal length; k is the vibration offset value, which can be positive, negative or zero; R 12 [ k ] is the reference frame and vibration frame signal in the X-axis or Y-axis direction h 1 and h 2 in vibration offset value k Cross-correlation value under ;

[0097] After obtaining the vibration offset value index k corresponding to the maximum cross-correlation value of the reference frame and the vibration frame in the X-axis and Y-axis directions, the peak correction method is used to correct the k, multiply the corrected offset in the corresponding direction by the scale factor to obtain the vibration magnitude in the X-axis and Y-axis directions;

[0098] The formula for calculating the scale factor in the X-axis or Y-axis direction is as follows:

[0099]

[0100] in, M is the scaling factor, L is the actual length of the composite vibration fringe pattern in the X-axis or Y-axis direction, a is the size of a single pixel in the imaging camera, r' It is the number of pixels on the camera sensor occupied by the side length of the vibration fringe image in the X-axis or Y-axis direction.

[0101] In this embodiment, for each frame of vibration fringe image, the fringe frequency is calculated using a frequency correction method; specifically:

[0102] The grayscale values of the vibration fringe image are subjected to Fourier transformation row by row and column by column to obtain the fringe frequency of each row and column in the vibration fringe image. The obtained fringe frequency of each row and column is corrected using the peak correction method to obtain multiple groups of fringe frequencies of the corrected vibration fringe image in the X-axis and Y-axis directions.

[0103] In this embodiment, the mathematical model based on frequency and object distance is used to calculate the single-dimensional vibration information of the mechanical structure in the Z-axis direction in the vibration fringe image of the current frame; specifically:

[0104] Construct a mathematical model of the fringe frequency and the distance between targets along the Z axis of the camera's optical axis:

[0105]

[0106] in, f I is the fringe frequency of the vibration fringe image on the sensor, D is the distance between the imaging camera and the composite vibration fringe pattern along the camera optical axis Z axis, F is the focal length of the imaging lens, a is the size of a single pixel in the imaging camera, p is the fringe period of the composite vibration-sensitive fringe pattern;

[0107] For the multiple groups of fringe frequencies of the corrected vibration fringe image in the X-axis and Y-axis directions, the median of the fringe frequencies of each column in the X-axis direction and the median of the fringe frequencies of each row in the Y-axis direction are averaged to obtain the frequency of the vibration fringe image on the sensor. f I , substitute into the mathematical model to calculate the object distance D , obtain the position information of the mechanical structure in the Z-axis direction.

[0108] In this embodiment, the step S33 is specifically as follows:

[0109] Convolve the vibration-fringe image:

[0110]

[0111] in R ( x , y ) is the Gabor filter at pixel position ( x , y ) output response at; I ( u , v ) is the vibration fringe image at position ( u , v ), The spatial frequency is f , direction is i Gabor filter kernel;

[0112] The two-dimensional calculation formula of the Gabor filter is as follows:

[0113]

[0114] in is a two-dimensional Gabor function, Represents the coordinates after rotation transformation:

[0115]

[0116] in, represents the Gabor filter wavelength, and ; represents the Gaussian standard deviation; Indicates the spatial aspect ratio; represents phase shift; represents an imaginary unit;

[0117] The Gabor filter wavelength Set to the fringe frequency in the vibration fringe image f I The inverse of the Gabor filter direction i The value range is set to 0-360°, and the step size is set to between 0.01° and 1°. According to the preset step size, all directions of the Gabor filter are traversed and the filter output response of each direction is calculated. The Gabor filter angle corresponding to the maximum output response is i , as the in-plane rotation angle of the mechanical structure in the Z-axis direction.

[0118] In this embodiment, the object distance is calculated based on the obtained frequency of each row and column fringe and the mathematical model of frequency and object distance, and the corrected object distance is obtained by data fitting; specifically:

[0119] In the X-axis direction, based on the obtained frequency of each fringe column, the corresponding object distance is calculated according to the mathematical model of frequency and object distance. D x1 , D x2 ,… D xc , and obtain the corrected object distance by least square fitting D' x1 , D' x2 ,… D' xc ;

[0120] In the Y-axis direction, based on the obtained frequency of each line of stripes, the corresponding object distance is calculated according to the mathematical model of frequency and object distance. D y1 , D y2 ,… D yr , and obtain the corrected object distance by least square fitting D' y1 , D' y2 ,… D' yr .

[0121] In this embodiment, the mathematical model derived between the composite vibration-sensitive fringe pattern and the mechanical structure vibration is used to obtain the out-of-plane rotation angle information of the mechanical structure according to the corrected object distance; specifically:

[0122] Out-of-plane rotation angle in the X-axis direction α Expressed as:

[0123]

[0124] in, D' x1 is the distance between the first row of fringe in the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction, D' xc is the distance between the last row of fringe in the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction; D is the distance between the imaging camera and the composite vibration fringe pattern along the camera optical axis Z-axis obtained in step S32; r x is the number of camera sensor pixels occupied by the composite vibration fringe pattern in the X-axis direction;

[0125] Out-of-plane rotation angle in the Y-axis direction β Expressed as:

[0126]

[0127] in, D' y1 is the distance between the first row of fringe in the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction, D' yr is the distance between the last row of the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction; r y is the number of camera sensor pixels occupied by the composite vibration fringe pattern in the Y-axis direction.

[0128] The present invention also proposes a mechanical vibration measurement device based on composite vibration fringes, which is implemented using any of the above-mentioned mechanical vibration measurement methods based on composite vibration fringes, and includes a composite vibration fringes target 3, a high-speed camera 1, and a computer 2;

[0129] The composite vibration fringe target is provided with a composite vibration fringe pattern, and the composite vibration fringe target is attached to the mechanical structure 4 to be tested, for obtaining vibration sensing information;

[0130] The high-speed camera 1 is responsible for capturing images of the composite vibration fringe pattern and transmitting the images to the computer 2;

[0131] The computer 2 extracts the vibration perception information of the target image and performs analysis and processing to obtain multi-dimensional vibration information of the mechanical structure.

[0132] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other manner. Any person skilled in the art may utilize the above-disclosed technical content to modify or modify the present invention into equivalent embodiments. However, any simple modifications, equivalent variations, and modifications to the above embodiments that do not depart from the technical content of the present invention and are based on the technical essence of the present invention remain within the scope of protection of the present invention.

Claims

1. A mechanical vibration measurement method based on composite vibration fringes, characterized in that: The following steps are involved: Step S1: Attach a composite vibration fringe pattern to the mechanical structure to be measured, and establish three-dimensional coordinates at the center point of the composite vibration fringe pattern. The camera optical axis perpendicular to the camera imaging surface is defined as the Z axis, the horizontal direction perpendicular to the imaging optical axis is defined as the X axis, and the vertical direction perpendicular to the imaging optical axis is defined as the Y axis, for sensing the vibration signal of the mechanical structure; Place a high-speed camera in a stable area, and set the focal length and aperture size according to the sensor size and shooting distance so that the composite vibration fringe pattern can be captured by the high-speed camera; Step S2: When the mechanical structure vibrates, the relative position of the composite vibration fringe pattern attached to the mechanical structure with respect to the imaging sensor of the high-speed camera also changes accordingly; a vibration fringe image with multi-dimensional vibration signals is acquired by the high-speed camera; Step S3: transmitting the collected vibration fringe signals to a computer, which decodes the vibration fringe signals and extracts multi-dimensional vibration information of the mechanical structure; specifically, the following steps are included: Step S31: For each frame of the vibration fringe image, the Pearson correlation coefficient between the grayscale value of the calculation area of the vibration fringe image and the grayscale value of the reference area in the X-axis and Y-axis directions is calculated, and a Pearson correlation coefficient curve is obtained; with the initial frame as the reference frame, the Pearson correlation coefficient curves of the reference frame and the vibration frame are cross-correlated in the X-axis and Y-axis directions at different vibration offset values, respectively, to obtain the vibration offset value index corresponding to the maximum cross-correlation value between the reference frame and the vibration frame in the X-axis and Y-axis directions, and the two-dimensional vibration information of the mechanical structure in the X-axis and Y-axis directions in the vibration fringe images of the reference frame and the vibration frame is calculated based on the obtained vibration offset value index; Step S32: For each frame of the vibration fringe image, the fringe frequency is calculated using a frequency correction method; and based on a mathematical model of frequency and object distance, the single-dimensional vibration information of the mechanical structure in the Z-axis direction in the current frame of the vibration fringe image is calculated; Step S33: according to the grayscale information of each frame of the vibration fringe image, a vibration fringe decoding strategy based on a Gabor filter is used to identify the in-plane rotation angle information of the mechanical structure; Step S34: Calculating the object distance based on the obtained row and column fringe frequencies and the mathematical model of frequency and object distance, and obtaining a corrected object distance through data fitting; deriving a mathematical model between the composite vibration-sensing fringe pattern and the vibration of the mechanical structure to obtain out-of-plane rotation angle information of the mechanical structure based on the corrected object distance; The composite vibration-sensitive stripe pattern is based on the sinusoidal stripes. By modulating the frequency and direction information of the sinusoidal stripes, a composite coded stripe pattern with the same frequency in different directions is formed, including a single sinusoidal stripe area in the middle and oblique stripe areas around it. The middle single sinusoidal stripe region includes four square single sinusoidal stripe regions, the edges of the four single sinusoidal stripe regions are respectively connected to the edges of the middle square gray block and form a cross with the gray block, and the stripe directions of the four single sinusoidal stripe regions are respectively parallel to the edges of the gray block; The surrounding oblique stripe area includes four square oblique stripe areas located in the blank areas around the middle single sinusoidal stripe area. The stripes of the four square oblique stripe areas have the same direction and are connected to the stripes of the single sinusoidal stripe area.

2. The mechanical vibration measurement method based on composite vibration fringes according to claim 1, characterized in that: The Pearson correlation coefficients between the grayscale values of the calculated region and the grayscale values of the reference region of the vibration fringe images in the X-axis and Y-axis directions are calculated respectively as follows: When calculating the Pearson correlation coefficient between the grayscale values of the calculation area and the grayscale values of the reference area of the vibration fringe image in the X-axis direction, the two single sinusoidal fringe areas with the stripes perpendicular to the X-axis and the square gray block in the middle are used as the reference areas, and the other areas except the reference area are used as the calculation areas; Take the grayscale value of any row in the reference area as the reference row grayscale value and calculate the Pearson correlation coefficient: where ρ r，i is the PCC correlation coefficient between the grayscale value of the reference row r and the grayscale value of the i-th row in the calculation area; I r is the grayscale value of the reference row, I i It is the gray value of the i-th row in the calculation area; cov(I r , I i ) is I r and I i covariance of σ r isI r The standard deviation, σ i isI i The standard deviation of I r，j is the grayscale value of the reference row r, column j, I i，j Calculate the grayscale value of the i-th row and j-th column in the area, where n is the number of columns; and I r and I i The mean of When calculating the Pearson correlation coefficient between the grayscale values of the calculation area and the reference area of the vibration fringe image in the Y-axis direction, the two single sinusoidal fringe areas with the stripes perpendicular to the Y-axis and the square gray block in the middle are used as the reference areas, and the other areas except the reference area are used as the calculation areas; Take any column of grayscale values in the reference area as the reference column grayscale value and calculate the Pearson correlation coefficient: where ρ c，j is the PCC correlation coefficient between the grayscale value of the reference column c and the grayscale value of the jth column of the calculation area; I c is the grayscale value of the reference column, I j is the gray value of the jth column in the calculation area; cov(I c , I j ) is I c and I j covariance of σ c isI c The standard deviation, σ j isI j The standard deviation of I c，i is the grayscale value of the i-th row in the reference column c, i j，i Calculate the grayscale value of the jth column and the ith row in the area, where m is the number of rows; and I c and I j The mean of .

3. The mechanical vibration measurement method based on composite vibration fringes according to claim 1, characterized in that: The method comprises performing cross-correlation calculation on the Pearson correlation coefficient curves of the reference frame and the vibration frame in the X-axis and Y-axis directions at different vibration offset values, obtaining the vibration offset value index corresponding to the maximum cross-correlation value between the reference frame and the vibration frame in the X-axis and Y-axis directions, and calculating the two-dimensional vibration information of the mechanical structure in the X-axis and Y-axis directions in the vibration fringe images of the reference frame and the vibration frame based on the obtained vibration offset value index; specifically, as follows: After Fourier transforming the obtained Pearson correlation coefficient curves in the X-axis and Y-axis directions, the Fourier transformed correlation coefficient curves are upsampled by zero padding in the frequency domain, and then the time domain information is returned by inverse Fourier transform to increase the pixel resolution of the vibration measurement. The cross-correlation calculation is performed using the processed Pearson correlation coefficient curves in the X-axis and Y-axis directions of the reference frame and the vibration frame: Where h1[s] is the Pearson correlation coefficient of the reference frame signal h1 in the X-axis or Y-axis direction at data point s, h2[s+k] is the Pearson correlation coefficient of the vibration frame signal h2 in the corresponding X-axis or Y-axis direction at data point s+k, S is the signal length; k is the vibration offset value, which can be positive, negative or zero; R 12 [k] is the cross-correlation value of the reference frame and vibration frame signals h1 and h2 in the X-axis or Y-axis direction at the vibration offset value k; After obtaining the vibration offset value index k corresponding to the maximum cross-correlation value between the reference frame and the vibration frame in the X-axis and Y-axis directions, k is corrected using the peak correction method. The corrected offset value in the corresponding direction is multiplied by the scale factor to obtain the vibration magnitude in the X-axis and Y-axis directions. The formula for calculating the scale factor in the X-axis or Y-axis direction is as follows: Where M is the scale factor, L is the actual length of the composite vibration fringe pattern in the X-axis or Y-axis direction, a is the size of a single pixel of the imaging camera, and r' is the number of pixels on the camera sensor occupied by the side length of the vibration fringe image in the X-axis or Y-axis direction.

4. The mechanical vibration measurement method based on composite vibration fringes according to claim 1, characterized in that: For each frame of vibration fringe image, the fringe frequency is calculated using a frequency correction method; specifically: The grayscale values of the vibration fringe image are subjected to Fourier transformation row by row and column by column to obtain the fringe frequency of each row and column in the vibration fringe image. The obtained fringe frequency of each row and column is corrected using the peak correction method to obtain multiple groups of fringe frequencies of the corrected vibration fringe image in the X-axis and Y-axis directions.

5. The mechanical vibration measurement method based on composite vibration fringes according to claim 4, characterized in that: The mathematical model based on frequency and object distance calculates the single-dimensional vibration information of the mechanical structure in the Z-axis direction in the vibration fringe image of the current frame; specifically: Construct a mathematical model of the fringe frequency and the distance between targets along the Z axis of the camera's optical axis: Among them, f I is the fringe frequency of the vibration fringe image on the sensor, D is the distance between the imaging camera and the composite vibration fringe pattern along the camera optical axis Z, F is the focal length of the imaging lens, a is the size of a single pixel of the imaging camera, and p is the fringe period of the composite vibration fringe pattern; For the multiple groups of fringe frequencies of the corrected vibration fringe image in the X-axis and Y-axis directions, the median of the fringe frequencies of each column in the X-axis direction and the median of the fringe frequencies of each row in the Y-axis direction are taken and averaged as the frequency f of the vibration fringe image on the sensor. I , substitute into the mathematical model to calculate the object distance D and obtain the position information of the mechanical structure in the Z-axis direction.

6. The mechanical vibration measurement method based on composite vibration fringes according to claim 1, characterized in that: The step S33 is specifically as follows: Convolve the vibration-fringe image: R(x,y)=∫∫I(u,v)g f，θ (x-u,y-v)dudv Where R(x,y) is the output response of the Gabor filter at the pixel position (x,y); I(u,v) is the grayscale value of the vibration fringe image at the position (u,v), g f，θ is a Gabor filter kernel with spatial frequency f and direction θ; The two-dimensional calculation formula of the Gabor filter is as follows: Where g(x, y) is a two-dimensional Gabor function, and (x′, y′) represents the coordinates after rotation transformation: Where λ represents the Gabor filter wavelength, and σ represents the Gaussian standard deviation; γ represents the spatial aspect ratio; ψ represents the phase shift; j′ represents the imaginary unit; The Gabor filter wavelength λ is set to the fringe frequency f in the vibration fringe image I The reciprocal of the value of the Gabor filter direction θ is set to 0-360°, and the step size is set to between 0.01° and 1°. All directions of the Gabor filter are traversed according to the preset step size and the filter output response in each direction is calculated. The Gabor filter angle θ corresponding to the maximum output response is used as the in-plane rotation angle of the mechanical structure in the Z-axis direction.

7. The mechanical vibration measurement method based on composite vibration fringes according to claim 5, characterized in that: The object distance is calculated based on the obtained frequency of each row and column fringe and the mathematical model of frequency and object distance, and the corrected object distance is obtained by data fitting; specifically: In the X-axis direction, based on the obtained frequency of each fringe column, the corresponding object distance D is calculated according to the mathematical model of frequency and object distance. x1 ,D x2 ,…D xc , and obtain the corrected object distance D' by least squares fitting x1 ,D' x2 ,…D' xc ; In the Y-axis direction, based on the obtained frequency of each line of stripes, the corresponding object distance D is calculated according to the mathematical model of frequency and object distance. y1 ,D y2 ,…D yr , and obtain the corrected object distance D' by least squares fitting y1 ,D' y2 ,…D' yr .

8. The mechanical vibration measurement method based on composite vibration fringes according to claim 7, characterized in that: The mathematical model derived between the composite vibration-sensing fringe pattern and the mechanical structure vibration is used to obtain the out-of-plane rotation angle information of the mechanical structure according to the corrected object distance; specifically: The out-of-plane rotation angle α in the X-axis direction is expressed as: Among them, D' x1 is the distance between the first row of fringe in the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction, D' xc is the distance between the last row of fringe patterns of the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction; D is the distance between the imaging camera and the composite vibration fringe pattern along the camera optical axis Z-axis obtained in step S32; r x is the number of camera sensor pixels occupied by the composite vibration fringe pattern in the X-axis direction; The out-of-plane rotation angle β in the Y-axis direction is expressed as: Among them, D' y1 is the distance between the first row of fringe in the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction, D' yr is the distance between the last row of the corrected composite vibration fringe pattern and the imaging camera in the Z-axis direction; r y is the number of camera sensor pixels occupied by the composite vibration fringe pattern in the Y-axis direction.

9. A mechanical vibration measurement device based on composite vibration fringes, characterized in that: The device is implemented by the mechanical vibration measurement method based on composite vibration fringe according to any one of claims 1 to 8, comprising a composite vibration fringe target, a high-speed camera, and a computer; The composite vibration-sensing stripe target is provided with a composite vibration-sensing stripe pattern, and the composite vibration-sensing stripe target is attached to the mechanical structure to be measured to obtain vibration sensing information; The high-speed camera is responsible for collecting images of the composite vibration-sensitive fringe pattern and transmitting the images to a computer; The computer extracts vibration perception information of the target image and performs analysis and processing to obtain multi-dimensional vibration information of the mechanical structure.

Images