Grating enhancement method for reducing deformation field measurement error under high noise condition

Abstract

The invention relates to the technical field of optical measurement and image processing, and discloses a grating enhancement method for reducing deformation field measurement errors under a high-noise condition, which comprises the following operation steps of: 1, performing Fourier transform on a grating image to obtain a frequency spectrum image; 2, measuring the distances fx and fyi of first-order dominant frequencies in the x direction and the y direction in the spectrum image; 3, a dot matrix mask is generated with the center of the frequency spectrum image as the zero point and with fx and fy as the distance, and the radius of each dot in the mask is R; and 4, extracting the data of the mask position in the spectrum image to obtain a periodic information spectrum. According to the grating enhancement method for reducing the deformation field measurement error under the high-noise condition, periodic information in a grating image can be effectively extracted and enhanced under the high-noise and complex background conditions, noise and background interference are remarkably suppressed, high-quality input is provided for subsequent deformation field measurement, and therefore the measurement accuracy is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of optical measurement and image processing, and particularly to a grating enhancement method for reducing the measurement error of a deformation field under high-noise conditions. Background Art

[0002] With the continuous development of optical measurement technology, deformation field measurement methods based on periodic structures (such as sampling moiré method, Fourier transform method, windowed Fourier transform method, geometric phase analysis method, etc.) are widely used in the fields of material mechanics, structural health monitoring, etc. Such methods usually apply a periodic grating structure on the surface of the object to be measured, and use image acquisition and processing means to obtain the field distribution information of the object during the stress or deformation process. As a deformation carrier, the periodic structure can effectively improve the measurement accuracy and spatial resolution, and thus has important value in practical engineering and scientific research.

[0003] Existing periodic structure deformation field measurement methods can obtain relatively accurate measurement results in an ideal or low-noise environment. However, in practical applications, the collected grating images are often affected by various high-noise factors such as complex background, uneven illumination, and sensor noise. These noise and background interferences will cause the periodic information in the image to be masked, thereby affecting the subsequent extraction and analysis of the deformation field and generating large measurement errors. For example, in a high-noise environment, deformation field measurement methods such as the sampling moiré method and the Fourier transform method are difficult to accurately separate the phase information of the periodic information, resulting in a decrease in the accuracy and robustness of the deformation field measurement.

[0004] In order to weaken the influence of complex background and high noise on the deformation field measurement, the prior art mainly uses means such as spatial filtering, frequency domain filtering, or image enhancement to preprocess the collected images. However, these methods often have deficiencies such as limited filtering effect, loss of periodic information, or strong dependence on parameters, and it is difficult to effectively suppress the interference brought by high noise and complex background while ensuring the original deformation integrity of the periodic information. Therefore, how to accurately extract the periodic information in the image under high-noise conditions and reduce the measurement error of the deformation field has become a technical problem that needs to be solved urgently in this field. Summary of the Invention

[0005] (I) Technical Problems to be Solved

[0006] In view of the deficiencies of the prior art, the present invention provides a grating enhancement method for reducing the measurement error of the deformation field under high-noise conditions. It has the advantages of extracting the frequency-domain dot matrix mask and enhancing the periodic information of the image, effectively suppressing noise and background interference, and improving the accuracy and robustness of the deformation field measurement. It solves the problem that the prior art mainly uses means such as spatial filtering, frequency-domain filtering or image enhancement to preprocess the acquired image. However, these methods often have deficiencies such as limited filtering effect, loss of periodic information or strong dependence on parameters, and it is difficult to effectively suppress the interference caused by high noise and complex background while ensuring the original deformation integrity of the periodic information.

[0007] (2) Technical solution

[0008] To achieve the above-mentioned purpose of extracting the frequency-domain dot matrix mask and enhancing the periodic information of the image, effectively suppressing noise and background interference, and improving the accuracy and robustness of the deformation field measurement, the present invention provides the following technical solution: A grating enhancement method for reducing the measurement error of the deformation field under high-noise conditions, including the following operation steps:

[0009] Step 1, perform a Fourier transform on the grating image to obtain a spectral image;

[0010] Step 2, measure the spacing f x between the first-order main frequencies in the x and y directions in the spectral image yi ;

[0011] Step 3, taking the center of the spectral image as the zero point, generate a dot matrix mask with f x and f y as the spacing, and the radius of each dot in the mask is R;

[0012] Step 4, extract the data at the mask position in the spectral image to obtain the periodic information spectrum;

[0013] Step 5, perform an inverse Fourier transform on the periodic information spectrum to obtain the image after grating enhancement.

[0014] Further, step 1 is specifically: perform a two-dimensional Fourier transform on the acquired grating image to obtain its spectral image. The spatial domain of the grating image is set as I(x, y), and the spectral image obtained by performing a two-dimensional Fourier transform on the grating image can be expressed as F(u, v):

[0015]

[0016] where w and h are the image sizes, representing the width and height of the image respectively.

[0017] Further, step 2 is specifically: measure the image pixel spacing f xWith f y , the first-order main frequency is the peak region closest to the center zero point in the frequency spectrum image, f x With f y are the pixel spacings between the center positions of the first-order main frequencies in the x and y directions and the center zero point, respectively, corresponding to the periodicity in the spatial domain.

[0018] Furthermore, the specific content of step three is as follows: Taking the center of the frequency spectrum image as the zero point, based on f x With f y generate a dot matrix mask. The function of the dot matrix mask is to filter out the frequency components related to the grating periodic structure in the frequency spectrum image. The radius R of the dot determines the filtering range, only retaining the frequency data related to the periodicity within a specific range and excluding other interfering frequency components. The mask M(u, v) can be defined as:

[0019]

[0020] where (u, v) are the coordinates in the frequency domain of the grating image, (u0, v0) is the center (zero order) in the Fourier spectrum, and k represents the order of the harmonic peak in the spectrum.

[0021] Furthermore, the specific content of step four is as follows: Extract the data at the mask positions in the frequency spectrum image to obtain the periodicity information frequency spectrum After the frequency domain of the image is extracted by the mask, it can be expressed as:

[0022]

[0023] Furthermore, the specific content of step five is as follows: Perform a two-dimensional inverse Fourier transform on the periodicity information frequency spectrum to obtain the image I ge (x, y) enhanced with periodicity information, which can be expressed as:

[0024]

[0025] Furthermore, the Fourier transform in step one is a mathematical transform that converts a time-domain (spatial-domain) signal to the frequency domain. For a grating image, through the Fourier transform, the image can be converted from the spatial domain to the frequency domain, enabling the frequency characteristics of the image to be displayed. Different structures and patterns in the image will have different manifestations in the frequency domain. For example, periodic structures will have specific frequency components in the frequency domain, facilitating subsequent analysis and processing.

[0026] Furthermore, extracting the data at the mask positions from the frequency spectrum image, the obtained periodicity information frequency spectrum only contains the frequency components related to the grating periodic structure, removing most of the noise and irrelevant frequency information, and extracting and focusing on the effective frequency information.

[0027] Further, the inverse Fourier transform described in Step Five is the inverse process of the Fourier transform. The periodic information spectrum extracted is subjected to the inverse Fourier transform to convert the frequency-domain information back to the spatial domain, obtaining the image enhanced by the grating. After the screening and processing of the frequency-domain information in the previous steps, the image after the inverse transform highlights the periodic structure of the grating and suppresses other interference factors, achieving the image enhancement effect.

[0028] (III) Beneficial Effects

[0029] Compared with the prior art, the present invention provides a grating enhancement method for reducing the measurement error of the deformation field under high-noise conditions, having the following beneficial effects:

[0030] 1. The grating enhancement method for reducing the measurement error of the deformation field under high-noise conditions can effectively extract and enhance the periodic information in the grating image under high-noise and complex background conditions, significantly suppressing noise and background interference, providing high-quality input for subsequent deformation field measurement, and thus significantly improving the measurement accuracy.

[0031] 2. The grating enhancement method for reducing the measurement error of the deformation field under high-noise conditions is applicable to the preprocessing steps of various optical deformation field measurement methods with periodic structures as deformation carriers, such as the sampling moiré method, the Fourier transform method, the windowed Fourier transform method, the geometric phase analysis method, etc., having a wide range of application prospects.

[0032] 3. The grating enhancement method for reducing the measurement error of the deformation field under high-noise conditions can effectively retain the deformation information of the periodic structure during the denoising process, improve the accuracy and noise resistance of the deformation field measurement, and has a high degree of automation, flexible parameter settings, and is convenient for practical engineering applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is the flowchart of the steps of the present invention;

[0034] Figure 2 is the detailed flowchart diagram of the actual application of the present invention;

[0035] Figure 3 is the processing effect diagram of the grating image with high noise and complex background of the present invention under different deformation degrees;

[0036] Figure 4 is the comparison of the x-direction root mean square error histograms of the phase differences obtained by the present invention and the traditional method;

[0037] Figure 5 is the comparison of the y-direction root mean square error histograms of the phase differences obtained by the present invention and the traditional method;

[0038] Figure 6Comparison of the full-field phase accuracy of the deformation measurement obtained by the present invention and the traditional method. Specific embodiments

[0039] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0040] Please refer to Figures 1-6 , a grating enhancement method for reducing the deformation field measurement error under high-noise conditions, including the following operation steps:

[0041] Step 1: Perform a Fourier transform on the grating image to obtain a spectral image;

[0042] Step 2: Measure the spacing f of the first-order main frequencies in the x and y directions in the spectral image x and f yi ;

[0043] Step 3: Taking the center of the spectral image as the zero point, generate a dot matrix mask with f x and f y as the spacing, and the radius of each dot in the mask is R;

[0044] Step 4: Extract the data at the mask position in the spectral image to obtain a periodic information spectrum;

[0045] Step 5: Perform an inverse Fourier transform on the periodic information spectrum to obtain the image after grating enhancement.

[0046] In the case implementation, Step 1 is specifically: perform a two-dimensional Fourier transform on the collected grating image to obtain its spectral image. The spatial domain of the grating image is set as I(x, y), and the spectral image obtained by performing a two-dimensional Fourier transform on the grating image can be expressed as F(u, v):

[0047]

[0048] where w and h are the image sizes, representing the width and height of the image respectively, and the noise information and periodic information can be distinguished by the brightness peaks in the spectral diagram;

[0049] Through numerical simulation, the measurement results and accuracy verification with the traditional method are carried out. A rectangular wave grating image is created through numerical simulation, such as Figure 2As shown in the simulated grating image (0% noise). The Peaks function in Matlab was used to apply theoretical deformation to this simulated image to obtain the deformed grating image. Subsequently, random noise and strip-shaped complex background noise were introduced to obtain a grating image with high noise and complex background, and its theoretical deformation was the same as that of the deformed grating image.

[0050] In the implementation of the case, step two is specifically: measuring the image pixel pitch f of the first-order main frequency in the x and y directions in the spectral image F(u, v) x and f y , the first-order main frequency is the peak region closest to the center zero point in the spectral image, and f x and f y are the pixel pitches between the center positions of the first-order main frequencies in the x and y directions and the center zero point, corresponding to the periodicity in the spatial domain;

[0051] As Figure 2 shown in the process, the spectral image of the grating with high noise and complex background is obtained through two-dimensional Fourier transform. The brightness peaks in the image contain the periodic information in the grating image with high noise and complex background. Peak extraction is performed through the above dot matrix mask to obtain the masked spectral image.

[0052] In the implementation of the case, step three is specifically: taking the center of the spectral image as the zero point, and generating a dot matrix mask according to f x and f y . The function of the dot matrix mask is to screen out the frequency components related to the periodic structure of the grating in the spectral image. The radius R of the dot determines the screening range, and only the frequency data related to the periodicity within a specific range is retained, excluding other interfering frequency components. The mask M(u, v) can be defined as:

[0053]

[0054] where (u, v) are the coordinates in the frequency domain of the grating image, (u0, v0) is the center (zero order) in the Fourier spectrum, and k represents the order of the harmonic peak in the spectrum;

[0055] Through the mask in the frequency domain, as Figure 2 shown in, which takes the zero order in the frequency domain as the center, and takes the image pixel pitches f of the first-order main frequencies in the x and y directions x and f y as the period, and R as the radius of a single mask circle, the periodic information in the spectral image is extracted. Subsequently, the inverse two-dimensional Fourier transform is performed on the masked spectral image to obtain the enhanced grating image, as Figure 2 shown in. The enhanced grating image effectively eliminates the previously introduced random noise and complex strip-shaped background noise.

[0056] In the case implementation, step four is specifically as follows: Extract the data at the mask position in the spectral image to obtain the periodic information spectrum. After the frequency domain of the image is extracted by the mask, it can be expressed as:

[0057]

[0058] By using the dot matrix mask to extract the data at the corresponding position in the spectral image, the periodic information spectrum is obtained. This process retains the periodic structural features in the grating image and filters out most of the noise and complex background information.

[0059] In the case implementation, step five is specifically as follows: Perform an inverse two-dimensional Fourier transform on the periodic information spectrum to obtain the image I ge (x, y) with enhanced periodic information, which can be expressed as:

[0060]

[0061] By using methods such as the sampling moiré method to analyze the image I ge (x, y) with enhanced periodic information, the obtained deformation field can greatly reduce the error caused by the complex background in the original image, making the sampling moiré method have stronger noise resistance.

[0062] In the case implementation, high-noise complex background grating images with different deformation degrees (Peak = 0%, 10%, 20%, 30%, 40%) were processed to obtain the Figure 3 group of images shown. This group of images shows the original simulated grating images with different deformation degrees, the complex background grating images after introducing noise, and the grating enhanced images processed by the present invention. In order to verify the suppression effect of the present invention on the measurement error after grating enhancement, the sampling moiré method was used to analyze the high-noise complex background grating images and the grating enhanced images, and the phase differences at different deformation degrees of the two groups of images were obtained. The sampling moiré method is a full-field deformation measurement method based on the grating image. By analyzing the phase of the grating and using the relationship between the phase difference and the displacement field and strain field, the full-field deformation is obtained. The root mean square error (RMSE) is used to evaluate this method and the phase error between it and the theoretical phase difference (theoretical deformation amount), where the root mean square error RMSE can be expressed as:

[0063]

[0064] In the case implementation, Figure 4 and Figure 5 show the histograms of the root mean square errors of the phase differences in the x and y directions obtained by the traditional method and after grating enhancement by the present invention. From Figure 4It can be seen that the root mean square error of the phase difference after grating enhancement by the present invention is greatly reduced, and the overall error suppression trend is obvious. Among them, when analyzing a grating image with a complex background of high noise with Peaks = 40% by the traditional method, the root mean square error reaches 20%. After grating enhancement by the present invention, the root mean square error is reduced to 2% when R = 6, verifying the accuracy of the present invention and the effectiveness of noise error suppression.

[0065] In the case implementation, the full-field phase error (compared with the phase difference after theoretical deformation) of the traditional method and the grating-enhanced one by the present invention was compared. As Figure 6 shown, among them, under the influence of a complex strip background, the error of the traditional method at the local part (high-noise position) reached more than 100%, while the full-field phase error after grating enhancement by the present invention was reduced to 5%, also demonstrating the effectiveness of the present invention in suppressing the error of a complex strip background with high noise.

[0066] In summary, the grating enhancement method for reducing the measurement error of the deformation field under high-noise conditions can effectively extract and enhance the periodic information in the grating image under high-noise and complex background conditions, significantly suppress noise and background interference, provide high-quality input for subsequent deformation field measurement, and thus significantly improve the measurement accuracy.

[0067] Moreover, the grating enhancement method for reducing the measurement error of the deformation field under high-noise conditions is applicable to the preprocessing steps of various optical deformation field measurement methods with a periodic structure as the deformation carrier, such as the sampling moiré method, the Fourier transform method, the windowed Fourier transform method, the geometric phase analysis method, etc. It has a wide application prospect, can effectively retain the deformation information of the periodic structure during the denoising process, improve the accuracy and noise resistance of the deformation field measurement, and has a high degree of automation, flexible parameter settings, and is convenient for practical engineering applications.

[0068] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the existence of additional identical elements in the process, method, article or device comprising the element.

[0069] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A grating enhancement method for reducing the deformation field measurement error under high-noise conditions, characterized in that: It includes the following operation steps: Step 1: Perform Fourier transform on the raster image to obtain a spectrum image; Step 2: Measure the spacing f between the first-order main frequencies in the x and y directions of the spectral image x and f yi ; Step 3: Take the center of the spectral image as the zero point, and generate a dot matrix mask with a spacing of f x and f y . Each dot in the mask has a radius of R; Step 4: Extract the data at the mask position in the spectrum image to obtain a periodic information spectrum; Step 5: Perform inverse Fourier transform on the periodic information spectrum to obtain the raster-enhanced image.

2. A grating enhancement method for reducing deformation field measurement errors under high-noise conditions according to claim 1, characterized in that: Specifically, Step 1 is as follows: Perform two-dimensional Fourier transform on the collected raster image to obtain its spectrum image. The spatial domain of the raster image is denoted as I(x, y), and the spectrum image obtained by performing two-dimensional Fourier transform on the raster image can be expressed as F(u, v): where w and h are the image sizes, representing the width and height of the image respectively.

3. A grating enhancement method for reducing deformation field measurement errors under high-noise conditions according to claim 1, characterized in that: The specific content described in Step 2 is as follows: Measure the image pixel pitch f of the first-order main frequency in the x and y directions in the spectral image F(u, v). x and f y , where the first-order main frequency is the peak region closest to the center zero point in the spectral image, and f x and f y are the pixel pitches between the center positions of the first-order main frequencies in the x and y directions and the center zero point, respectively, corresponding to the periodicity in the spatial domain.

4. A grating enhancement method for reducing deformation field measurement errors under high-noise conditions according to claim 1, characterized in that: Specifically, as described in Step 3: taking the center of the spectral image as the zero point, according to f x and f y a dot matrix mask is generated. The function of the dot matrix mask is to filter out the frequency components related to the grating periodic structure in the spectral image. The radius R of the dot determines the filtering range, and only the frequency data related to periodicity within a specific range is retained, excluding other interfering frequency components. The mask M(u, v) can be defined as: where (u, v) are the coordinates in the frequency domain of the raster image, (u0, v0) is the center (zero order) in the Fourier spectrum, and k represents the order of the harmonic peak in the spectrum.

5. A grating enhancement method for reducing deformation field measurement errors under high-noise conditions according to claim 1, characterized in that: The specific content described in Step 4 is as follows: Extract the data at the mask position in the spectral image to obtain the periodic information spectrum After the mask extraction, the frequency domain of the image can be expressed as:

6. A grating enhancement method for reducing the deformation field measurement error under high-noise conditions according to claim 1, characterized in that: Step 5 specifically includes: performing an inverse two-dimensional Fourier transform on the periodic information spectrum to obtain an image I ge (x, y) after enhancing the periodic information, which can be expressed as:

7. A grating enhancement method for reducing the deformation field measurement error under high-noise conditions according to claim 1, characterized in that: The Fourier transform in Step 1 is a mathematical transform that converts a time-domain (spatial-domain) signal to a frequency domain. For a raster image, through Fourier transform, the image can be converted from the spatial domain to the frequency domain, enabling the display of the frequency characteristics of the image. Different structures and patterns in the image will have different manifestations in the frequency domain. For example, periodic structures will have specific frequency components in the frequency domain, facilitating subsequent analysis and processing.

8. A grating enhancement method for reducing deformation field measurement errors under high-noise conditions according to claim 1, characterized in that: The extraction of the data at the mask position from the spectrum image in Step 4 results in a periodic information spectrum that only contains the frequency components related to the periodic structure of the raster, removing most of the noise and irrelevant frequency information, and extracting and focusing on the effective frequency information.

9. A grating enhancement method for reducing deformation field measurement errors under high-noise conditions according to claim 1, characterized in that: The inverse Fourier transform in Step 5 is the inverse process of the Fourier transform. Performing inverse Fourier transform on the extracted periodic information spectrum converts the frequency-domain information back to the spatial domain to obtain the raster-enhanced image. After the screening and processing of the frequency-domain information in the previous steps, the image after inverse transform highlights the periodic structure of the raster and suppresses other interference factors, achieving the image enhancement effect.

Images