A phase analysis-based optical interference fringe displacement prediction method and system

Abstract

The application provides a phase analysis-based optical interference fringe displacement prediction method and system, and belongs to the cross technical field of image processing and optical interference measurement. The method comprises the following steps: obtaining two optical interference fringe images to be detected, i.e., a reference fringe image and a target fringe image, wherein the target fringe image is an image obtained by translating the reference fringe image; pre-processing the obtained fringe images, eliminating redundant information and interference, and strengthening the fringe phase features; converting the fringe images from a spatial domain to a frequency domain through phase correlation calculation, extracting pure phase information related to displacement, and inversely transforming back to the spatial domain to obtain a phase correlation matrix; and completing fringe displacement detection based on the phase correlation matrix. The application solves the common problems of the prior art, such as complicated process, large amount of calculation, weak anti-interference and insufficient precision, and significantly improves the real-time performance, stability and sensitivity of methane concentration monitoring.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of image processing and optical interferometry, specifically involving optical interferometric fringe displacement prediction technology based on phase analysis, which is applicable to optical interferometric fringe displacement measurement scenarios with translation relationships. Background Technology

[0002] In the complex working environment of underground coal mines, methane accumulation can easily lead to safety accidents such as explosions and poisoning. Accurate and real-time monitoring of methane concentration is a crucial aspect of ensuring safe mine production. Optical interferometry methane detectors, based on the Michelson interferometer principle, utilize the absorption characteristics of methane molecules to specific wavelengths of light. By analyzing the displacement changes of interference fringes, methane concentration is detected. With its high precision and non-contact measurement advantages, it has become the mainstream equipment in the field of methane detection in coal mines. As the demand for industrial automation continues to upgrade, traditional manual detection methods are no longer suitable for the efficient and accurate monitoring requirements on-site. Digital upgrades of optical interferometry methane detectors based on image recognition technology have become the core development direction in this field.

[0003] In the prior art, CN 119207624 A discloses a "multi-step preprocessing + multi-dimensional feature extraction" optical interference fringe processing scheme. This scheme first converts the interference fringe image into a grayscale image and completes the black fringe line spectrum recognition. Then, it completes the scale contour fitting and horizontal correction through morphological dilation and Hough transform. Subsequently, it extracts the spectral peak and phase information through Fourier transform, normalization, and smoothing to achieve methane concentration detection. This scheme has shortcomings in practical applications: First, the processing flow is redundant. It requires multi-step preprocessing and feature extraction to enhance the fringe amplitude contrast, which involves a large amount of computation and high system computing power requirements, resulting in insufficient fringe displacement processing efficiency and failing to meet the needs of real-time on-site monitoring. Second, its anti-interference ability and detection accuracy are limited. Its feature extraction is highly dependent on the recognition of spectral amplitude peaks, which is easily affected by downhole lighting fluctuations and dust interference, causing deviations in core feature extraction. Moreover, it does not have a sub-pixel level precision optimization design, and cannot effectively capture minute fringe displacements.

[0004] CN 114113093 A discloses a digital processing scheme for optical interference fringes using "image optimization + virtual scale assistance." This scheme magnifies and clarifies the images of interference fringes and scales, while constructing a virtual scale to assist in identifying the alignment relationship between the fringes and the scale lines, thereby achieving digital readings of methane concentration. However, this scheme has several technical shortcomings: First, it lacks a core anti-interference processing mechanism for the complex downhole environment and does not possess the ability to actively extract fringe displacement features, making it a passive processing approach with insufficient operational stability. Second, the automation level of the entire detection process is low, requiring manual assistance to determine the alignment relationship between the fringes and the scale lines. Furthermore, it lacks specialized detection methods for minute displacements, making it difficult to capture the micro-displacements of fringes corresponding to low concentrations of methane, thus limiting the sensitivity of methane concentration detection.

[0005] In summary, both of the aforementioned existing technical solutions, as the mainstream technical paths for the digital upgrade of optical interferometry methane detection, share common problems that urgently need to be addressed: First, the detection process is cumbersome and computationally intensive, making it difficult to meet the application requirements of real-time downhole monitoring; second, they lack anti-interference mechanisms adapted to complex downhole working conditions, are easily affected by environmental factors, and have insufficient detection stability; and third, they have not established a sub-pixel-level precision optimization system, making it impossible to accurately capture minute stripe displacements and limiting the detection sensitivity for low-concentration methane. Summary of the Invention

[0006] In view of this, the purpose of this invention is to provide a method and system for predicting optical interference fringe displacement based on phase analysis. Based on existing optical interference fringe displacement measurement technology, this method and system abandon the multi-step redundant processing architecture and adopt a collaborative technical solution of "phase correlation method as the core + normalized cross power spectrum anti-interference + sub-pixel accuracy optimization" to directly locate fringe displacement through frequency domain phase features, thereby achieving efficient, stable and high-precision detection of interference fringe displacement.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for predicting optical interference fringe displacement based on phase analysis, the method specifically includes the following steps: S1. Obtain two optical interference fringe images to be detected: reference fringe image. and target stripe image The target stripe image is the reference stripe image after translation. The obtained image; S2. Preprocess the acquired stripe image to eliminate redundant information and interference, and enhance the stripe phase features. S3. Transform the stripe image from the spatial domain to the frequency domain by phase correlation calculation, extract the pure phase information related to displacement, and then inversely transform it back to the spatial domain to obtain the phase correlation matrix. S4. Perform stripe displacement detection based on the phase correlation matrix.

[0009] Furthermore, in step S4, the fringe displacement detection based on the phase correlation matrix specifically includes: S41. Pixel-level displacement is obtained through peak positioning; S42. Subpixel-level displacement is obtained by optimizing the parabolic fitting algorithm. S43, Outputs precise stripe displacement results.

[0010] Furthermore, in step S2, the preprocessing of the acquired stripe image specifically includes: S21. Assume the size of the acquired stripe image is... ,in Image width, The image height has coordinates that satisfy... The preprocessing process consists of two steps, which successively achieve interference elimination and feature enhancement. S22. Grayscale processing: For the baseline stripe image and target stripe image The image is converted to a single-channel grayscale image using a weighted average method to eliminate the interference of color information on phase extraction. The formula is as follows:

[0011]

[0012] in, These are the pixel values ​​of the red, green, and blue channels of the image, respectively. After processing, a striped grayscale image is obtained. and ; S23. Noise Suppression Processing: Gaussian filtering is used to smooth high-frequency noise such as dust and lighting fluctuations that may exist in the striped image, balancing noise suppression with stripe detail preservation, as detailed below: Gaussian filter kernel definition:

[0013] Among them, the filter kernel parameters ; Filtering operation implementation:

[0014]

[0015] in This represents the convolution operation, which removes high-frequency noise and preserves the core phase distribution features of the stripes.

[0016] Furthermore, in step S3, based on the displacement property of Fourier transform, the translation of spatial fringes is transformed into a linear phase change in the frequency domain. Pure phase information is extracted through three steps of computation, specifically including: S31. Two-dimensional Fast Fourier Transform (2D-FFT): This transforms the preprocessed fringe grayscale image from the spatial domain to the frequency domain, establishing the relationship between the spatial domain shift and the frequency domain phase. The formula is as follows:

[0017]

[0018] in, Frequency domain coordinates , j The imaginary unit is used; according to the Fourier transform displacement theorem, the fringe spatial domain is translated. The corresponding frequency domain phase satisfies: ; S32. Normalized cross-power spectrum calculation: To eliminate interference from illumination fluctuations, fringe contrast changes, and other factors, only pure phase information related to displacement is retained. The calculation process is as follows: Complex conjugate operation: obtaining complex conjugate :

[0019] in, for The real part of a complex number, for The imaginary part of a complex number; Normalized cross-power spectrum formula:

[0020] in Representing the modulus of a complex number; combining the frequency domain shift theorem, after normalization, we can obtain:

[0021] This result completely eliminates amplitude interference and retains only pure phase information, thus improving the algorithm's anti-interference capability. S33. Inverse Fourier Transform (IFFT2): Transforms the normalized cross-power spectrum from the frequency domain back to the spatial domain to obtain the phase correlation matrix, as shown in the following formula:

[0022] Phase correlation matrix In stripe displacement Sharp peaks will appear at the corresponding coordinates, providing a core basis for displacement detection.

[0023] Furthermore, in step S4, a two-level detection architecture of "pixel-level rapid positioning + sub-pixel-level precise optimization" is adopted. First, a rough displacement is obtained, and then the detection accuracy is improved. Specifically, this includes: S41, Pixel-level stripe displacement detection: 1) Peak location: Find the peak position with the largest magnitude in the phase correlation matrix. This peak is generated by the phase matching corresponding to the fringe translation, as shown in the following formula:

[0024] in The magnitudes of the phase correlation matrix elements and the peak coordinates are given. Pixel-level clues directly corresponding to stripe displacement; 2) Pixel-level displacement calculation: Since the image array index starts from 1, coordinate correction is performed on the peak position, as shown in the following formula:

[0025]

[0026] 3) Boundary Correction Processing: Considering the periodicity of the Fourier transform, if the peak position is close to the image boundary, cyclic boundary correction is required to avoid displacement calculation errors. x Direction correction: when hour, ;when hour, ; y Direction correction: when hour, ;when hour,

[0027] S42. Subpixel-level fringe displacement optimization: To accurately capture minute fringe displacements while maintaining good real-time performance, a parabolic fitting subpixel localization method is adopted. It is assumed that the phase correlation matrix can be approximated by a parabola in the peak neighborhood. The implementation steps are as follows: 1) Neighborhood point selection: Consistent with the Gaussian fitting method, peak points are selected. and three neighboring areas and ; 2) Subpixel offset calculation: x direction:

[0028] y direction:

[0029] Final subpixel displacement:

[0030] .

[0031] The present invention also provides an optical interference fringe displacement prediction system based on phase analysis, which employs the method described above.

[0032] Furthermore, the system is integrated into an optical interferometer methane detector to achieve real-time, high-precision detection of methane concentration.

[0033] The beneficial effects of this invention are as follows: Compared to existing technologies, this invention, through a synergistic technical solution of "phase correlation method as the core + normalized cross-power spectrum anti-interference + sub-pixel accuracy optimization," accurately solves the core pain point of optical interference fringe displacement measurement in methane detection scenarios in underground coal mines. Specific beneficial effects are as follows: The processing efficiency is significantly improved, which can fully adapt to the needs of real-time monitoring in wells. This invention abandons the redundant architecture of multi-step preprocessing and multi-dimensional feature extraction in the existing technology, and adopts a simplified processing flow of "two-dimensional FFT transformation - conjugate calculation - normalized cross power spectrum - inverse FFT", which reduces the algorithm time complexity to (O(Nlog N)) (N is the number of image pixels), greatly improving the real-time response capability of stripe displacement detection.

[0034] The anti-interference capability is greatly enhanced, effectively improving the detection stability under complex working conditions. This invention relies on frequency domain conjugate product and normalized cross-power spectrum processing to completely shield various interferences caused by amplitude changes, retaining only pure phase features directly related to fringe displacement, without the need for additional interference suppression modules or supporting auxiliary hardware.

[0035] This invention achieves a breakthrough in detection accuracy, significantly enhancing the detection sensitivity for low-concentration methane. By introducing a parabolic fitting algorithm in the peak region of the phase correlation matrix, it successfully achieves sub-pixel accuracy detection, effectively overcoming the technical limitations of existing technologies in detecting low-concentration methane.

[0036] In summary, this invention solves the common problems of existing optical interferometric fringe displacement measurement technologies by synergistically optimizing detection efficiency, anti-interference capability under complex working conditions, and measurement accuracy across multiple dimensions. It provides complete, efficient, and practical technical support for the digital upgrade of optical interferometric methane detectors. The application of this invention can significantly improve the real-time performance, stability, and sensitivity of methane concentration monitoring in coal mines, effectively strengthening the technical foundation for underground gas control in coal mines. It has crucial engineering practical significance and industry promotion value for ensuring safe mine production and preventing gas accidents.

[0037] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0038] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein: Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0039] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0040] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0041] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0042] This invention addresses the shortcomings of existing optical interferometric fringe displacement measurement technology in methane detection in underground coal mines by providing an optimized displacement prediction method. This method can: 1) solve the problems of redundant processes and high computational load in existing technologies, reduce computing power requirements, improve processing efficiency, and meet real-time monitoring requirements; 2) overcome the defects of weak anti-interference ability, adapt to complex underground working conditions, reduce the impact of environmental interference, and enhance detection stability; 3) compensate for the shortcomings of insufficient sub-pixel level accuracy, improve the ability to capture small fringe displacements, and increase the sensitivity of methane concentration detection.

[0043] Figure 1 The flowchart of the method of the present invention is shown in the figure. The main steps of the method provided by the present invention include: firstly, acquiring two optical interference fringe images to be detected (reference fringe images). Target stripe image ,in for After translation The obtained stripe image is then preprocessed to eliminate redundant information and interference, and enhance the stripe phase features. Next, the stripe image is transformed from the spatial domain to the frequency domain through phase correlation calculation to extract pure phase information related to displacement, and then inversely transformed back to the spatial domain to obtain the phase correlation matrix. Finally, stripe displacement detection is completed based on the phase correlation matrix. First, the pixel-level displacement is obtained through peak localization, and then the sub-pixel-level displacement is obtained through parabolic fitting algorithm optimization, and finally, accurate stripe displacement results are output.

[0044] Specifically, it includes: (1) The image preprocessing method is as follows: Assume the size of the acquired stripe image is... ( W Image width, H (Image height), coordinates satisfying The preprocessing process consists of two steps, which successively achieve interference cancellation and feature enhancement: ① Grayscale processing: for the reference stripe image and target stripe image The image is converted to a single-channel grayscale image using a weighted average method to eliminate the interference of color information on phase extraction. The formula is as follows:

[0045]

[0046] in, These are the pixel values ​​of the red, green, and blue channels of the image, respectively. After processing, a striped grayscale image is obtained. and .

[0047] ② Noise Suppression Processing: Gaussian filtering is used to smooth the striped image, addressing potential high-frequency noise such as dust and lighting fluctuations. This balances noise suppression with stripe detail preservation, as detailed below: Gaussian filter kernel definition:

[0048] Among them, the filter kernel parameters .

[0049] Filtering operation implementation:

[0050]

[0051] in This represents the convolution operation, which removes high-frequency noise and preserves the core phase distribution features of the stripes.

[0052] (2) The phase correlation calculation method is as follows: Based on the displacement property of Fourier transform, the translation of spatial fringes is transformed into a linear phase change in the frequency domain. Pure phase information is extracted through three-step operation. The specific implementation process is as follows: ① Two-dimensional Fast Fourier Transform (2D-FFT): This transforms the preprocessed fringe grayscale image from the spatial domain to the frequency domain, establishing the relationship between the spatial domain shift and the frequency domain phase. The formula is as follows:

[0053]

[0054] in, Frequency domain coordinates , j The unit is the imaginary unit. According to the Fourier transform displacement theorem, the fringe spatial domain is translated... The corresponding frequency domain phase satisfies:

[0055] ② Normalized cross-power spectrum calculation: To eliminate interference from illumination fluctuations, fringe contrast changes, and other factors, only pure phase information related to displacement is retained. The calculation process is as follows: Complex conjugate operation: obtaining complex conjugate :

[0056] in, for The real part of a complex number, for The imaginary part of a complex number.

[0057] Normalized cross-power spectrum formula:

[0058] in Let represent the modulus of the complex number. Combining this with the frequency-domain shift theorem, after normalization, we obtain:

[0059] This result completely eliminates amplitude interference and retains only pure phase information, thus improving the algorithm's anti-interference capability.

[0060] ③ Inverse Fourier Transform (IFFT2): Transforms the normalized cross-power spectrum from the frequency domain back to the spatial domain to obtain the phase correlation matrix, as shown in the following formula:

[0061] Phase correlation matrix In stripe displacement Sharp peaks will appear at the corresponding coordinates, providing a core basis for displacement detection.

[0062] (3) The stripe displacement detection method is as follows: a two-level detection architecture of "pixel-level fast positioning + sub-pixel-level precise optimization" is adopted. First, a rough displacement amount is obtained, and then the detection accuracy is improved. The specific implementation process is as follows: ①Pixel-level stripe displacement detection Peak location: Find the peak position with the largest magnitude in the phase correlation matrix. This peak is generated by the phase matching corresponding to the fringe translation, as shown in the following formula:

[0063] in The magnitudes of the phase correlation matrix elements and the peak coordinates are given. Pixel-level clues directly corresponding to stripe displacement.

[0064] Pixel-level displacement calculation: Since the image array index starts from 1, coordinate correction is performed on the peak position, as shown in the following formula:

[0065]

[0066] Boundary correction processing: Considering the periodicity of the Fourier transform, if the peak position is close to the image boundary, cyclic boundary correction is required to avoid displacement calculation errors. x Direction correction: when hour, ;when hour, ; y Direction correction: when hour, ;when hour, .

[0067] ② Subpixel-level stripe displacement optimization To accurately capture minute stripe displacements while maintaining good real-time performance, a parabolic fitting subpixel positioning method is employed.

[0068] Assuming the phase correlation matrix can be approximated by a parabola in the peak neighborhood, the implementation steps are as follows: Neighborhood point selection: Consistent with the Gaussian fitting method, peak points are selected. and three neighboring areas and ; Subpixel offset calculation: x direction:

[0069] y direction:

[0070] Final subpixel displacement:

[0071] .

[0072] In summary, the technical solution of this invention solves the common problems of existing technologies, such as cumbersome processes, large computational load, weak anti-interference, and insufficient accuracy. It significantly improves the real-time performance, stability, and sensitivity of methane concentration monitoring, provides complete technical support for the digital upgrade of optical interferometric methane detectors, and reduces application costs and operational barriers. The main innovative technical points of this invention are as follows: 1) Abandoning multi-step preprocessing and multi-dimensional feature extraction, the stripe displacement is located by the peak value of the phase correlation matrix through the process of "two-dimensional FFT transformation - conjugate calculation - normalized cross power spectrum - inverse FFT". This simplifies the process, reduces the computing power requirement (time complexity O(NlogN)), avoids preprocessing distortion, improves detection reliability and response speed, and meets the needs of real-time monitoring in downhole.

[0073] 2) By using frequency domain conjugate product and normalization processing, the influence of amplitude changes is shielded and anti-interference phase characteristics are preserved. No additional interference suppression module or auxiliary hardware is required. It effectively resists light fluctuations and dust interference, reduces feature extraction deviation, lowers equipment costs, and improves detection stability under complex working conditions.

[0074] 3) A parabolic fitting algorithm is introduced in the peak region of the phase correlation matrix to balance real-time performance and accuracy. It accurately captures minute stripe displacements, breaks through the limitations of pixel-level accuracy, adapts to the needs of downhole real-time monitoring and low-concentration methane detection, and reduces displacement errors and concentration conversion deviations.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for predicting optical interference fringe displacement based on phase analysis, characterized in that, The method specifically includes the following steps: S1. Obtain two optical interference fringe images to be detected: reference fringe image. and target stripe image The target stripe image is the reference stripe image after translation. The obtained image; S2. Preprocess the acquired stripe image to eliminate redundant information and interference, and enhance the stripe phase features. S3. Transform the stripe image from the spatial domain to the frequency domain by phase correlation calculation, extract the pure phase information related to displacement, and then inversely transform it back to the spatial domain to obtain the phase correlation matrix. S4. Perform stripe displacement detection based on the phase correlation matrix.

2. The optical interference fringe displacement prediction method based on phase analysis according to claim 1, characterized in that, In step S4, the fringe displacement detection based on the phase correlation matrix specifically includes: S41. Pixel-level displacement is obtained through peak positioning; S42. Subpixel-level displacement is obtained by optimizing the parabolic fitting algorithm. S43, Outputs precise stripe displacement results.

3. The optical interference fringe displacement prediction method based on phase analysis according to claim 2, characterized in that, In step S2, the preprocessing of the acquired stripe image specifically includes: S21. Assume the size of the acquired stripe image is... ,in Image width, The image height has coordinates that satisfy... The preprocessing process consists of two steps, which successively achieve interference elimination and feature enhancement. S22. Grayscale processing: For the baseline stripe image and target stripe image The image is converted to a single-channel grayscale image using a weighted average method to eliminate the interference of color information on phase extraction. The formula is as follows: in, These are the pixel values ​​of the red, green, and blue channels of the image, respectively. After processing, a striped grayscale image is obtained. and ; S23. Noise Suppression Processing: Gaussian filtering is used to smooth high-frequency noise in the striped image, balancing noise suppression with stripe detail preservation, as detailed below: Gaussian filter kernel definition: Among them, the filter kernel parameters ; Filtering operation implementation: in This represents the convolution operation, which removes high-frequency noise and preserves the core phase distribution features of the stripes.

4. The optical interference fringe displacement prediction method based on phase analysis according to claim 3, characterized in that, In step S3, based on the displacement property of Fourier transform, the translation of spatial fringes is transformed into a linear phase change in the frequency domain. Pure phase information is extracted through three steps of operation, specifically including: S31. Two-dimensional Fast Fourier Transform: This transforms the preprocessed fringe grayscale image from the spatial domain to the frequency domain, establishing the relationship between the spatial domain shift and the frequency domain phase. The formula is as follows: in, Frequency domain coordinates , j The imaginary unit is used; according to the Fourier transform displacement theorem, the fringe spatial domain is translated. The corresponding frequency domain phase satisfies: ； S32. Normalized cross-power spectrum calculation: To eliminate interference from illumination fluctuations and fringe contrast changes, only pure phase information related to displacement is retained. The calculation process is as follows: Complex conjugate operation: obtaining complex conjugate : in, for The real part of a complex number, for The imaginary part of a complex number; Normalized cross-power spectrum formula: in Representing the modulus of a complex number; combining the frequency domain shift theorem, after normalization, we can obtain: This result completely eliminates amplitude interference and retains only pure phase information, thus improving the algorithm's anti-interference capability. S33. Inverse Fourier Transform (IFFT2): Transforms the normalized cross-power spectrum from the frequency domain back to the spatial domain to obtain the phase correlation matrix, as shown in the following formula: Phase correlation matrix In stripe displacement Sharp peaks will appear at the corresponding coordinates, providing a core basis for displacement detection.

5. The optical interference fringe displacement prediction method based on phase analysis according to claim 4, characterized in that, Step S4 specifically includes: S41, Pixel-level stripe displacement detection: 1) Peak location: Find the peak position with the largest magnitude in the phase correlation matrix. This peak is generated by the phase matching corresponding to the fringe translation, as shown in the following formula: in The magnitudes of the phase correlation matrix elements and the peak coordinates are given. Pixel-level clues directly corresponding to stripe displacement; 2) Pixel-level displacement calculation: Since the image array index starts from 1, coordinate correction is performed on the peak position, as shown in the following formula: 3) Boundary Correction Processing: Considering the periodicity of the Fourier transform, if the peak position is close to the image boundary, cyclic boundary correction is required to avoid displacement calculation errors. x Direction correction: when hour, ;when hour, ; y Direction correction: when hour, ;when hour, S42. Subpixel-level fringe displacement optimization: To accurately capture minute fringe displacements while maintaining good real-time performance, a parabolic fitting subpixel localization method is adopted. It is assumed that the phase correlation matrix can be approximated by a parabola in the peak neighborhood. The implementation steps are as follows: 1) Neighborhood point selection: Consistent with the Gaussian fitting method, peak points are selected. and three neighboring areas and ; 2) Subpixel offset calculation: x direction: y direction: Final subpixel displacement: 。 6. A system for predicting optical interferometric fringe displacement based on phase analysis, characterized in that, The system employs the method as described in any one of claims 1-5.

7. The optical interferometric fringe displacement prediction system based on phase analysis according to claim 6, characterized in that, The system is integrated into an optical interferometer methane detector, enabling real-time, high-precision detection of methane concentration.

Images