Computer system and method based on a model of reconstruction of axial strain field
Patent Information
- Application Number
- CN202610751794.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-28
- Publication Date
- 2026-09-11
- Estimated Expiration
- 2046-05-28
AI Technical Summary
[0004]然而，光矢量法在应变估计过程中存在一个本质瓶颈，即复矢量相位的数值微分过程会将噪声纳入计算从而累积测量误差，并最终导致错误的应变结果
[0056] The computer system based on the axial strain field reconstruction model disclosed herein first acquires the interference spectrum inside the material before and after mechanical loading using phase-sensitive OCT; then, it constructs a complex phase signal using the phase difference signal; and introduces a two-dimensional Gaussian window function into the frequency domain analysis of the original signal to filter and judge the signal coherence among multiple frequency components. This not only achieves effective separation of the true phase signal and random phase noise, but also further constructs a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient through mathematical derivation; finally, the axial strain field is reconstructed using the linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient. Since it does not require pixel-by-pixel estimation of the phase signal in the depth axis direction as in the traditional optical vector method, accurate strain reconstruction under high-noise conditions can be achieved without numerical differentiation operations. At the same time, the computer system provided by this disclosure also has high sensitivity to deformation gradient features, and the strain resolution is improved by an order of magnitude compared with the traditional optical vector method, which has broad application prospects.
Smart Images
Figure CN122282464B_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of computer technology, and in particular to a computer system and method based on an axial strain field reconstruction model. Background Technology
[0002] Optical coherence tomography (OCT) is an internal full-field imaging technique with micrometer-level spatial resolution. It achieves cross-sectional imaging by real-time detection of the interference spectrum, which characterizes the internal information of materials. It has been widely applied in industrial non-destructive testing fields such as coating thickness measurement, micro-defect identification, and multilayer component evaluation. Subsequently, phase-sensitive OCT was further proposed for dynamic mechanical analysis under sample loading conditions. It utilizes the difference in interference phase at different monitoring times to characterize the instantaneous mechanical properties of materials. During loading, the material wavelength changes due to changes in refractive index; a half-wavelength optical path difference corresponds to a phase change of 2π radians. Therefore, this technique possesses nanometer-level displacement detection sensitivity and is widely used as a precision monitoring method in key production processes such as polymer curing manufacturing, optical fiber drawing, and composite material layup.
[0003] In practical measurements, axial strain serves as a crucial indicator for evaluating the quality of industrial materials. It effectively reveals localized stress concentrations caused by microscopic defects or process deficiencies, providing key criteria for material production and process optimization. Phase-sensitive OCT can quantitatively estimate axial strain by measuring the gradient changes between phase fringes. This estimation primarily involves two methods: the time-domain phase dewinding method and the complex-domain phase numerical differentiation method. The former struggles to effectively unwind the phase information wound in the [-π, π] region into a continuous signal due to phase noise, thus making it unsuitable for industrial scenarios with varying noise levels. The latter, by mapping the phase signal to the complex domain without phase dewinding, has become the more mainstream strain estimation method. This method primarily uses a vector smoothing window to mean-filter the phase noise and then performs a depth-axis numerical differentiation operation on the vector phase to effectively estimate the phase gradient and obtain the strain distribution; therefore, this method is also known as the optical vector method.
[0004] However, the optical vector method suffers from a fundamental bottleneck in strain estimation: the numerical differentiation of the complex vector phase incorporates noise into the calculation, accumulating measurement errors and ultimately leading to erroneous strain results. While this method can enhance phase noise suppression in the preprocessing stage by increasing the spatial sliding filter window, this inevitably blurs the boundary characteristics of the phase gradient, reduces the strain resolution of the local analysis window region, and consequently masks the non-uniform mechanical characteristics of material deformation, resulting in distorted strain estimation. Therefore, the optical vector method typically relies on the experience of technicians to adjust the window size, thus balancing noise suppression with accurate strain measurement. Consequently, it is essential to find a high-resolution strain reconstruction method that does not require numerical differentiation, effectively suppresses the interference of phase noise on strain estimation, and achieves high-precision strain resolution measurement.
[0005] Therefore, there is an urgent need for an axial strain field reconstruction model that can fundamentally overcome the above limitations, accurately estimate strain under high noise conditions, and significantly improve measurement accuracy in terms of strain resolution. Summary of the Invention
[0006] The purpose of this disclosure is to provide a computer system and method based on an axial strain field reconstruction model, which is used to solve at least one technical problem in the prior art.
[0007] The technical solution disclosed herein is:
[0008] A computer system based on an axial strain field reconstruction model, comprising:
[0009] The acquisition module is used to obtain the interference spectrum inside the material before and after mechanical loading via phase-sensitive OCT.
[0010] The preprocessing module interacts with the acquisition module to obtain the temporal phase signal in the interference spectrum; and uses the temporal phase signal to obtain the phase difference signal to map the optical path change caused by the internal deformation of the material before and after mechanical loading.
[0011] The separation module interacts with the preprocessing module to construct a complex phase signal using the phase difference signal and analyzes the complex phase signal using a two-dimensional Gaussian window function to separate the effective phase difference signal and phase noise.
[0012] The processing module interacts with the separation module to process the effective phase difference signal in a preset local area. A linear approximation is performed on the surface to construct a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient; and the relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient is obtained based on the linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient.
[0013] The output module interacts with the processing module to obtain the axial strain inside the material based on the relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient.
[0014] The relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient is as follows:
[0015] At any point in the center of the local region, the frequency signal corresponding to the peak position of the spectral energy at that point is the phase gradient at that point.
[0016] The effective phase difference signal is then applied in a preset local region. A linear approximation is applied to construct a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient, including:
[0017] The linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient is expressed as follows:
[0018]
[0019] in, This represents the two-dimensional windowed Fourier transform spectrum of a local window. express Fourier transform; Represents the window function. = Indicates the center is A two-dimensional Gaussian function; | | is a frequency in an instantaneous period =( The Gaussian bell-shaped function that reaches its maximum value at () ; |∙| denotes the modulo function; Represents the imaginary unit; Indicates the center point of the local window The phase difference signal value at that location; Indicates the lateral relative offset; Indicates the relative axial offset; This represents the first-order partial derivative of the phase difference signal in the transverse direction; This represents the first-order partial derivative of the phase difference signal along the axis; and These represent the spectral signal in the horizontal direction. Shaft and axial direction Instantaneous frequency on the axis.
[0020] Obtaining the axial strain inside the material includes:
[0021] In the center of the local area At this point, the two-dimensional windowed Fourier transform spectrum of the local window.
[0022] | The frequency signal corresponding to the peak position represents the phase gradient at that peak point. At this time, axial strain Represented as:
[0023] ;
[0024] in, Represents the horizontal coordinates of the phase difference image; Represents the axial coordinates of the phase difference image; Indicates the center wavelength of the light source; Indicates the phase difference signal; Indicates the axial phase gradient; Indicates the coordinates of the center of the local window; Indicates the instantaneous axial frequency; Indicates the transverse instantaneous frequency; Indicates the retrieval of spectral energy Maximum Instantaneous frequency of the shaft.
[0025] Obtaining a phase difference signal using the timing phase signal includes:
[0026] ;
[0027] in, and These represent the horizontal coordinates and axial coordinates of the phase difference image, respectively. and These represent the recorded phases before and after material deformation.
[0028] The process of constructing a complex phase signal using the phase difference signal and analyzing the phase signal using a two-dimensional Gaussian window function to accurately separate the effective phase difference signal and phase noise includes:
[0029] A complex phase signal is constructed from the phase difference signal. :
[0030] Introducing a two-dimensional Gaussian window function Analyze the spectral energy of the local phase region;
[0031] For complex phase signals Perform a windowed Fourier transform, which is represented as:
[0032] ;
[0033] in, This represents the two-dimensional windowed Fourier transform spectrum of a local window. Indicates the center is A two-dimensional Gaussian function; and These represent the spectral signals at... shaft and Instantaneous frequency on the axis; Represents the horizontal coordinates of the phase difference image; Represents the axial coordinates of the phase difference image; Represents the imaginary unit; This is the Fourier transform kernel function.
[0034] The introduction of a two-dimensional Gaussian window function The spectral energy of the local phase region is analyzed, including:
[0035] The two-dimensional Gaussian window function The spectral energy used for local phase region analysis is expressed as: ;
[0036] in, These represent the Gaussian window function at... and Pixel expansion in direction; These are the normalization coefficients; It is a Gaussian distribution kernel function.
[0037] The complex phase signal is constructed using the phase difference signal. ,include:
[0038] The complex phase signal is constructed from the phase difference signal using the following model. :
[0039] ;
[0040] in, Represents an exponential function; Represents the imaginary unit; and These represent the horizontal coordinates and axial coordinates of the phase difference image, respectively. This represents the phase difference signal.
[0041] The effective phase difference signal is then applied in a preset local region. The surface is approximated linearly, including:
[0042] Within the local region, the phase difference signal exist The surface is approximated linearly; where, at the point... After applying a first-order Taylor expansion, Represented as:
[0043] ;
[0044] in, This represents the two-dimensional windowed Fourier transform spectrum of a local window. Represents an exponential function; Represents the imaginary unit; Represents the coordinates of the center point of a local area; Indicates the phase difference at the center point; Indicates the lateral phase gradient; Indicates the axial phase gradient; Represents a two-dimensional Gaussian window function; Indicates the instantaneous axial frequency; Indicates the transverse instantaneous frequency;
[0045] And, order , To construct a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient; where, Indicates the lateral relative offset; Indicates the relative axial offset; and These represent the horizontal coordinates and axial coordinates of the phase difference image, respectively.
[0046] A high-resolution strain reconstruction method, based on the aforementioned computer system using an axial strain field reconstruction model, includes:
[0047] Interference spectra inside the material before and after mechanical loading were acquired using phase-sensitive OCT.
[0048] The temporal phase signal in the interference spectrum is obtained; and the phase difference signal is obtained using the temporal phase signal to map the optical path change caused by the internal deformation of the material before and after mechanical loading.
[0049] A complex phase signal is constructed using the phase difference signal, and the phase signal is analyzed using a two-dimensional Gaussian window function to accurately separate the effective phase difference signal and phase noise.
[0050] The effective phase difference signal is applied in a preset local region. A linear approximation is applied to construct a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient;
[0051] The relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient is obtained based on the linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient.
[0052] The axial strain inside the material is obtained based on the relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient.
[0053] The relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient is as follows:
[0054] At any point in the center of the local region, the frequency signal corresponding to the peak position of the spectral energy at that point is the phase gradient at that point.
[0055] The beneficial effects of this disclosure include at least the following:
[0056] The computer system based on the axial strain field reconstruction model disclosed herein first acquires the interference spectrum inside the material before and after mechanical loading using phase-sensitive OCT; then, it constructs a complex phase signal using the phase difference signal; and introduces a two-dimensional Gaussian window function into the frequency domain analysis of the original signal to filter and judge the signal coherence among multiple frequency components. This not only achieves effective separation of the true phase signal and random phase noise, but also further constructs a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient through mathematical derivation; finally, the axial strain field is reconstructed using the linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient. Since it does not require pixel-by-pixel estimation of the phase signal in the depth axis direction as in the traditional optical vector method, accurate strain reconstruction under high-noise conditions can be achieved without numerical differentiation operations. At the same time, the computer system provided by this disclosure also has high sensitivity to deformation gradient features, and the strain resolution is improved by an order of magnitude compared with the traditional optical vector method, which has broad application prospects. Attached Figure Description
[0057] Figure 1 This is a system block diagram of a computer system based on an axial strain field reconstruction model.
[0058] Figure 2 This is a flowchart of the high-resolution strain reconstruction method described in this disclosure;
[0059] Figure 3 This is a schematic diagram of the phase signal measurement process using a phase-sensitive OCT.
[0060] Figure 4 This is a schematic diagram of the windowed Fourier transform process of the original phase signal using a two-dimensional Gaussian function.
[0061] Figure 5 A schematic diagram for axial strain reconstruction using the instantaneous frequency of the local maximum spectral energy;
[0062] Figure 6 This is a comparison diagram of the strain reconstruction process between the high-resolution strain reconstruction method described in this disclosure and the traditional optical vector method;
[0063] Figure 7The graph shows a comparison of the measurement results of the high-resolution strain reconstruction method and the optical vector method described in this disclosure at t=3s, t=5s, t=7s, and t=9s, respectively.
[0064] Figure 8 This is a comparison chart of the strain resolution of the high-resolution strain reconstruction method and the optical vector method described in this disclosure. Detailed Implementation
[0065] The technical solution of this disclosure will be further described below with reference to the accompanying drawings. Specific Implementation Example 1:
[0067] This disclosure provides an embodiment:
[0068] like Figure 1 A computer system based on an axial strain field reconstruction model includes: an acquisition module 100, a preprocessing module 200, a separation module 300, a processing module 400, and an output module 500. The acquisition module 100 acquires the interference spectrum inside the material before and after mechanical loading using phase-sensitive OCT. The preprocessing module 200 interacts with the acquisition module 100 to acquire the temporal phase signal in the interference spectrum and uses the temporal phase signal to obtain a phase difference signal to map the optical path change caused by deformation inside the material before and after mechanical loading. The separation module 300 interacts with the preprocessing module 200 to construct a complex phase signal using the phase difference signal and analyze the complex phase signal using a two-dimensional Gaussian window function to separate the effective phase difference signal and phase noise. The processing module 400 interacts with the separation module 300 to output the effective phase difference signal within a preset local analysis window region. The surface is linearly approximated to construct a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient; and the relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient is obtained based on the linear mapping model; the output module 500 interacts with the processing module 400 to obtain the axial strain inside the material based on the relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient. Specific Implementation Example 2:
[0070] This disclosure provides another embodiment:
[0071] like Figure 2 Based on specific embodiment 1, this disclosure provides a high-resolution strain reconstruction method based on an axial strain field reconstruction model, which can perform strain measurement without numerical differentiation, accurately achieve strain estimation under high noise conditions, and improve measurement accuracy by 10 times in strain resolution, including the following steps:
[0072] S1: The interference spectrum inside the material before and after mechanical loading is acquired by phase-sensitive OCT. After demodulating and extracting the temporal phase signal, the phase difference is obtained through differential operation. The phase difference signal, which characterizes the optical path change caused by internal deformation of the material, can be expressed as:
[0073] (1)
[0074] in, and These represent the horizontal coordinates and axial coordinates of the phase difference image, respectively. and These represent the recorded phases before and after material deformation, respectively. The complete phase difference signal acquisition process is as follows: Figure 3 As shown.
[0075] S2: Using the phase difference signal to construct a complex phase signal, it can be expressed as:
[0076] (2)
[0077] in, Represents an exponential function. It represents the imaginary unit.
[0078] like Figure 4 As shown, the two-dimensional Gaussian window function The spectral energy introduced for local phase region analysis can be expressed as:
[0079] (3)
[0080] in, and These represent the Gaussian window function at... and Pixel expansion in direction. For The windowed Fourier transform can be represented as:
[0081] (4)
[0082] in, express Transform spectrum, Indicates the center is Two-dimensional Gaussian function, and These represent the spectral signals at... shaft and The instantaneous frequency on the axis. After the original phase undergoes a windowed Fourier transform, its signal analysis is limited to a certain range. In the frequency domain, the coherence of the periodic phase is effectively enhanced and the interference effect of random noise is weakened, accurately separating the effective signal and phase noise.
[0083] S3: For a defined local analysis window region on the image, the phase difference signal... It is possible The surface is approximated linearly, and its points are... The expression can be represented by a first-order Taylor expansion as follows:
[0084] (5)
[0085] in, Let represent the derivative function. Then formula (4) can be rewritten as:
[0086] (6)
[0087] make , Substituting it into formula (6) yields the following:
[0088] (7)
[0089] in, express The Fourier transform of . From formula (7), it can be seen that | | is a frequency in an instantaneous period =( The Gaussian bell-shaped function that reaches its maximum value at () is given by |•|, where |•| represents the modulo function.
[0090] S4: According to formula (7), in the center of the local analysis window area At that point, spectral energy
[0091] | The frequency signal corresponding to the peak position represents the phase gradient at that point. At this time, the axial strain It can be represented as:
[0092] (8)
[0093] in, Indicates the retrieval of spectral energy Maximum shaft instantaneous frequency The retrieval process is as follows: Figure 5 As shown.
[0094] Therefore, the proposed high-resolution strain reconstruction method only needs to perform each Calculate its local spectrum Then find the one that makes | Maximum instantaneous frequency The phase gradient at that point can then be obtained. This allows for the reconstruction of the axial strain field.
[0095] Verification process:
[0096] During material deformation, displacement scattering of internal microparticles generates random interference signals, which manifest as phase noise in phase contrast images. Due to the complexity of industrial environments, materials typically exhibit non-uniform deformation distribution under loading conditions, leading to multiple frequency variations in the noisy phase signal. Therefore, signal identification capability under low signal-to-noise ratio conditions is crucial. To demonstrate the advantages of the local maximum spectral energy method, its strain reconstruction process compared to the traditional optical vector method is presented. Figure 6 middle.
[0097] from Figure 6 It can be concluded that the traditional optical vector method requires two operations: smoothing and denoising the phase difference signal, and numerical differentiation. In the denoising process of the smoothing window, this method essentially treats the signal in the local analysis window region as a single frequency component, ignoring the role of other frequency information and making it difficult to obtain a high signal-to-noise ratio filtered phase difference signal. Furthermore, this method also estimates the phase gradient of adjacent pixels through numerical differentiation, which easily incorporates phase noise into the calculation, leading to a continuous accumulation of measurement errors.
[0098] The high-resolution strain reconstruction method disclosed herein does not require pixel-by-pixel estimation of the phase signal along the depth axis as in the traditional optical vector method. Therefore, this method can achieve accurate strain reconstruction under high-noise conditions without numerical differentiation operations. Furthermore, this method exhibits high sensitivity to deformation gradient features. Actual measurement experiments have verified that this method possesses micrometer-level strain resolution, an order of magnitude improvement over the optical vector method—that is, 10 times higher accuracy than the optical vector method in strain calculation.
[0099] Specific examples:
[0100] To verify the effectiveness of the proposed novel method, this step involves simultaneously reconstructing the strain of a polymer sample undergoing the same deformation process using both optical vector method and high-resolution strain reconstruction method. The sample is an epoxy resin film with a centrally located silicone rubber cube. Because the two materials have different elastic moduli, their strain distributions at the same time point also differ significantly. Figure 7As shown in figures a-1 to a-4, the phase difference distribution images of the sample at 3 seconds, 5 seconds, 7 seconds, and 9 seconds are presented, respectively. It can be observed that the phase fringe density continuously increases over time, and the fringe density in the central silicone rubber region is always greater than that in other areas covered by the epoxy resin material, indicating that the sample is undergoing overall deformation, and that the two materials exhibit different degrees of deformation. To illustrate the influence of the deformation process on the strain reconstruction effect, Figure 7 Figures b-1 to b-4 show the noise distribution corresponding to the aforementioned phase difference signal. It can be observed that as the deformation intensifies, the noise distribution also continuously increases and spreads, causing the image signal-to-noise ratio (SNR) to decrease from 14.7 dB to 12.6 dB.
[0101] Since traditional optical vector methods primarily estimate phase gradients through numerical differentiation between pixels, such noise variations can easily cause measurement errors and lead to strain reconstruction failure. Figure 7 As shown in c-1 to c-4, the optical vector method can accurately estimate the strain distribution of the sample under low-noise conditions starting at t=3s, where silicone rubber exhibits a greater strain than epoxy resin. At t=5s, the increase in phase noise significantly affects the strain measurement of this method, especially by introducing inaccurate random signals in the central region. By t=7s, this distortion has gradually spread to the entire bottom of the sample, and by t=9s, the entire sample fails to reconstruct properly, making it impossible to identify the strain distribution.
[0102] However, when the high-resolution strain reconstruction method described in this disclosure is used to perform actual measurements of phase difference signals under various noise conditions, such as... Figure 7 As shown in d-1 to d-4, the method described in this disclosure is not significantly affected by noise distribution. It not only has a significantly higher image signal-to-noise ratio than the light vector method, but also accurately and clearly characterizes the strain distribution of the two materials and their time-varying strain accumulation effect. Therefore, the high-resolution strain reconstruction method proposed in this disclosure can accurately and effectively realize strain reconstruction of different materials under multi-deformation and multi-noise environments.
[0103] To further verify the measurement performance of the method provided in this disclosure, the measurement resolution of the high-resolution strain reconstruction method described in this disclosure is compared with that of the traditional optical vector method after strain reconstruction, i.e., the so-called strain resolution. This index can be characterized by calculating the full width at half maximum (FWHM) of the point spread function (PSF), which can be obtained by differentiating the step response at the gradient boundary location along the z-axis.
[0104] like Figure 8As shown in (a), after the light vector method smooths the phase noise using a 10×10 pixel spatial window, the strain distribution within the yellow dashed box region exhibits significant gradient characteristics. The strain resolution at this location can be measured to be 67.58 μm from its PSF curve. Subsequently, as... Figure 8 In (b) and (c), larger smooth windows of 15×15 and 20×20 were used for strain reconstruction, yielding strain resolutions of 74.12 μm and 79.57 μm, respectively. This phenomenon indicates that as the window size increases, the ability of the light vector method to identify gradient features decreases, and may lead to erroneous strain boundary reconstruction results, as shown in the white box. Figure 8 In (d), the local maximum spectral energy method does not require the constant adjustment of the window size as in the optical vector method to seek a trade-off between noise suppression and accurate strain estimation. The reconstructed strain distribution exhibits clear boundary features and a high image signal-to-noise ratio. More notably, the method achieves an FWHM of 6.54 μm, a sensitivity improvement of an order of magnitude compared to the optical vector method, reaching micrometer-level strain resolution—that is, 10 times higher accuracy than the optical vector method in strain calculation. This demonstrates that the proposed local maximum spectral energy method possesses superior strain reconstruction capabilities. Specific Implementation Example 3:
[0106] This disclosure also provides an embodiment:
[0107] An electronic device includes: a storage medium and a processing unit; wherein the storage medium is used to store a computer program, and the processing unit exchanges data with the storage medium for executing the computer program during strain reconstruction to perform the steps of the method as described in Specific Embodiment 2. Specific Implementation Example 4:
[0109] A computer-readable storage medium storing a computer program; when the computer program is run, it performs the steps of the method as described in Specific Embodiment 2.
[0110] In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wireline, optical fiber, RF, etc., or any suitable combination thereof.
[0111] The above disclosure only discloses a few specific implementation scenarios. However, this disclosure is not limited to these. Any variations that can be conceived by those skilled in the art should fall within the protection scope of this disclosure.
Claims
1. A computer system based on an axial strain field reconstruction model, characterized in that, include: The acquisition module is used to obtain the interference spectrum inside the material before and after mechanical loading via phase-sensitive OCT. The preprocessing module interacts with the acquisition module to obtain the temporal phase signal in the interference spectrum; and uses the temporal phase signal to obtain the phase difference signal to map the optical path change caused by the internal deformation of the material before and after mechanical loading. The separation module interacts with the preprocessing module to construct a complex phase signal using the phase difference signal and analyzes the complex phase signal using a two-dimensional Gaussian window function to separate the effective phase difference signal and phase noise. The processing module interacts with the separation module to process the effective phase difference signal in a preset local area. A linear approximation is performed on the surface to construct a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient; and the relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient is obtained based on the linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient. The output module interacts with the processing module to obtain the axial strain inside the material based on the relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient. The relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient is as follows: At any point in the center of the local region, the frequency signal corresponding to the peak position of the spectral energy at that point is the phase gradient at that point. The effective phase difference signal is then applied in a preset local region. A linear approximation is applied to construct a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient, including: The linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient is expressed as follows: ； in, Represents the two-dimensional windowed Fourier transform spectrum of a local window; express Fourier transform; Represents the window function. Indicates the center is A two-dimensional Gaussian function; It is a frequency in an instant The Gaussian bell-shaped function that reaches its maximum value at a given point; Represents the modulo function; Represents the imaginary unit; Indicates the center point of the local window The phase difference signal value at that location; Indicates the lateral relative offset; Indicates the relative axial offset; This represents the first-order partial derivative of the phase difference signal in the transverse direction; This represents the first-order partial derivative of the phase difference signal along the axis; and These represent the spectral signal in the horizontal direction. Shaft and axial direction Instantaneous frequency on the axis; Obtaining the axial strain inside the material includes: In the center of the local area At this point, the two-dimensional windowed Fourier transform spectrum of the local window. The frequency signal corresponding to the peak position represents the phase gradient at that peak point. At this time, axial strain Represented as: ； in, Represents the horizontal coordinates of the phase difference image; Represents the axial coordinates of the phase difference image; Indicates the center wavelength of the light source; Indicates the phase difference signal; Indicates the axial phase gradient; Indicates the coordinates of the center of the local window; Indicates the instantaneous axial frequency; Indicates the transverse instantaneous frequency; Indicates the retrieval of spectral energy Maximum Instantaneous frequency of the shaft.
2. The computer system based on the axial strain field reconstruction model according to claim 1, characterized in that, Obtaining a phase difference signal using the timing phase signal includes: ； in, and These represent the horizontal coordinates and axial coordinates of the phase difference image, respectively. and These represent the recorded phases before and after material deformation.
3. The computer system based on the axial strain field reconstruction model according to claim 1, characterized in that, The process of constructing a complex phase signal using the phase difference signal and analyzing the phase signal using a two-dimensional Gaussian window function to accurately separate the effective phase difference signal and phase noise includes: A complex phase signal is constructed using the phase difference signal. : Introducing a two-dimensional Gaussian window function Analyze the spectral energy of the local phase region; For complex phase signals Perform a windowed Fourier transform, which is represented as: ； in, Represents the two-dimensional windowed Fourier transform spectrum of a local window; Indicates the center is A two-dimensional Gaussian function; and These represent the spectral signals at... shaft and Instantaneous frequency on the axis; Represents the horizontal coordinates of the phase difference image; Represents the axial coordinates of the phase difference image; Represents the imaginary unit; This is the Fourier transform kernel function.
4. The computer system based on the axial strain field reconstruction model according to claim 3, characterized in that, The introduction of a two-dimensional Gaussian window function The spectral energy of the local phase region is analyzed, including: The two-dimensional Gaussian window function The spectral energy used for local phase region analysis is expressed as: ; in, These represent the Gaussian window function at... and Pixel expansion in direction; These are the normalization coefficients; It is a Gaussian distribution kernel function.
5. The computer system based on the axial strain field reconstruction model according to claim 3, characterized in that, The complex phase signal is constructed using the phase difference signal. ,include: The following model is used to construct a complex phase signal from the phase difference signal. : ； in, Represents an exponential function; Represents the imaginary unit; and These represent the horizontal coordinates and axial coordinates of the phase difference image, respectively. This represents the phase difference signal.
6. The computer system based on the axial strain field reconstruction model according to claim 1, characterized in that, The effective phase difference signal is then applied in a preset local region. The surface is approximated linearly, including: Within the local region, the phase difference signal exist The surface is approximated linearly; where, at the point... After applying a first-order Taylor expansion, Represented as: ； in, Represents the two-dimensional windowed Fourier transform spectrum of a local window; Represents an exponential function; Represents the imaginary unit; Represents the coordinates of the center point of a local area; Indicates the phase difference at the center point; Indicates the lateral phase gradient; Indicates the axial phase gradient; Represents a two-dimensional Gaussian window function; Indicates the instantaneous axial frequency; Indicates the transverse instantaneous frequency; And, order , To construct a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient; where, Indicates the lateral relative offset; Indicates the relative axial offset; and These represent the horizontal coordinates and axial coordinates of the phase difference image, respectively.
7. A high-resolution strain reconstruction method, based on the computer system according to any one of claims 1-6, characterized in that, include: Interference spectra inside the material before and after mechanical loading were acquired using phase-sensitive OCT. Obtain the temporal phase signal from the interference spectrum; The phase difference signal is obtained by using the time-series phase signal to map the optical path change caused by the internal deformation of the material before and after mechanical loading. A complex phase signal is constructed using the phase difference signal, and the phase signal is analyzed using a two-dimensional Gaussian window function to accurately separate the effective phase difference signal and phase noise. The effective phase difference signal is applied in a preset local region. A linear approximation is applied to construct a linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient; The relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient is obtained based on the linear mapping model between the instantaneous frequency of the maximum spectral energy and the phase gradient. The axial strain inside the material is obtained based on the relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient. The relationship between the instantaneous frequency of the maximum spectral energy and the phase gradient is as follows: At any point in the center of the local region, the frequency signal corresponding to the peak position of the spectral energy at that point is the phase gradient at that point.