Laser-ultrasound signal processing method for dual-wave mixing interferometer

Abstract

The application relates to a laser ultrasonic signal processing method for a double-wave mixing interferometer, which comprises the following steps: acquiring a laser ultrasonic time domain signal received by the double-wave mixing interferometer; performing discrete wavelet decomposition on the laser ultrasonic time domain signal to obtain layer detail coefficients and approximation coefficients; determining adaptive threshold values for the layer detail coefficients based on unbiased risk estimation, and performing smoothing processing on the layer detail coefficients; performing inverse wavelet transformation and reconstruction on the smoothed detail coefficients and the reserved approximation coefficients to obtain a reconstructed signal; constructing a digital band-pass filter with maximum flat characteristics of a passband based on the frequency spectrum distribution of a target ultrasonic characteristic wave to be extracted; and extracting a target ultrasonic characteristic signal by using the digital band-pass filter. The adaptive soft threshold algorithm of wavelet decomposition and unbiased risk estimation can completely eliminate the signal artifacts and additional oscillations of a hard threshold value, and perfectly reserve the weak and key transient wave peaks and amplitude characteristics of an ultrasonic longitudinal wave.

Description

Technical Field

[0001] This invention belongs to the field of laser ultrasound and signal processing technology, specifically relating to a laser ultrasound signal processing method for a dual-wave mixing interferometer. Background Technology

[0002] Laser ultrasonic nondestructive testing (DUT) technology has been widely used in the fields of material internal defect detection and high-resolution imaging due to its advantages such as non-contact and wide bandwidth. In advanced laser ultrasonic testing systems, dual-wave mixing interferometers are often used to detect minute displacements on the material surface caused by ultrasonic wave propagation. However, the raw ultrasonic time-domain signal received by the interferometer is usually extremely complex, containing not only useful high-frequency volume waves (such as longitudinal waves) carrying information about internal material defects, but also a large number of low-frequency interference waves generated by laser excitation (such as surface waves and transverse waves), as well as high-frequency noise introduced by the system. From a physical acoustic perspective, high-frequency ultrasonic signals are easily attenuated when propagating inside materials, and the energy of the reflected echoes from minute defects they carry is often very weak and easily drowned out by high-intensity low-frequency signals. To achieve high-resolution detection and accurate quantitative analysis (such as assessing the size and depth of defects) of defects such as micropores inside materials, it is necessary to accurately extract high-frequency longitudinal wave signals in specific frequency bands from the complex raw time-domain signal. This requires the design and introduction of appropriate signal filters to effectively eliminate interference from low-frequency wave trains and suppress high-frequency noise. In current signal processing practices, Chebyshev filters, elliptic filters, or Bessel filters are commonly used for band separation and signal processing. However, these conventional filters have significant technical limitations when processing extremely sensitive laser-ultrasonic defect signals: First, the amplitude information of the laser ultrasonic signal (such as the amplitude of the defect echo) is the core basis for quantitative defect analysis. Some traditional filters (such as Chebyshev and elliptic filters) have obvious amplitude fluctuations (i.e., "ripples") in the passband, which will "distort" the amplitude characteristics of the useful signal, causing distortion and loss of the key defect signal energy characteristics, and thus leading to serious errors in the assessment of defect size and depth.

[0003] Secondly, the energy of low-frequency interference (such as surface waves and shear waves) and high-frequency noise typically exhibits a monotonic variation with frequency. If the stopband attenuation characteristics of the filter are not smooth enough, the interference signal at a specific frequency may be unexpectedly amplified due to the ripple effect, making it impossible to achieve stable suppression of out-of-passband interference. Therefore, the field of nondestructive testing urgently needs a laser-ultrasonic signal processing method specifically for dual-wave mixing interferometers. This method must be able to accurately filter out low-frequency shear wave and surface wave interference and high-frequency noise, while ensuring excellent passband flatness and smooth attenuation characteristics to ensure zero distortion of the amplitude and energy information of the characteristic echoes of minute defects, thereby providing a solid and reliable data foundation for subsequent high-precision quantitative defect analysis. Summary of the Invention

[0004] To address the problems raised in the background art, a laser ultrasonic signal processing method for a dual-wave mixing interferometer is provided in a first aspect of the present invention, comprising: Acquire the laser-ultrasound time-domain signal received by a dual-wave mixing interferometer; perform discrete wavelet decomposition on the laser-ultrasound time-domain signal to obtain detail coefficients and approximation coefficients at each level; Based on unbiased risk estimation, an adaptive threshold is determined for the detail coefficients of each layer, and the detail coefficients of each layer are smoothed according to the soft threshold rule; the smoothed detail coefficients and the retained approximation coefficients are then subjected to inverse wavelet transform and reconstruction to obtain the reconstructed signal; Based on the spectral distribution of the ultrasonic feature waves of the target to be extracted, a digital bandpass filter with the maximum flatness characteristic in the passband is constructed. The reconstructed signal is subjected to frequency band extraction using the digital bandpass filter to obtain the target ultrasonic feature signal.

[0005] In some embodiments of the present invention, determining the adaptive threshold corresponding to each level of detail coefficients based on unbiased risk estimation includes: constructing an unbiased risk estimation function corresponding to a candidate threshold for each level of detail coefficients; searching for a candidate threshold that minimizes the unbiased risk estimation function; and using the searched candidate threshold as the optimal denoising threshold for that level of detail coefficients.

[0006] In some embodiments of the present invention, the smoothing of the detail coefficients of each layer according to the soft threshold rule includes: setting the detail coefficients whose absolute value is less than or equal to the corresponding adaptive threshold to zero, and shrinking the detail coefficients whose absolute value is greater than the corresponding adaptive threshold to zero while keeping the sign unchanged.

[0007] In some embodiments of the present invention, constructing a digital bandpass filter with maximum flatness in the passband based on the spectral distribution of the ultrasonic feature wave of the target to be extracted includes: Construct the prototype transfer function of Butterworth low-pass filter based on the cutoff angular frequency and the filter order; Based on the lower and upper cutoff frequencies of the ultrasonic feature wave of the target to be extracted, the Butterworth low-pass prototype transfer function is mapped to an analog bandpass transfer function. The analog bandpass transfer function is converted into a digital transfer function in the z-domain to obtain the digital bandpass filter.

[0008] Furthermore, the step of mapping the Butterworth low-pass prototype transfer function to an analog bandpass transfer function includes: determining the center angular frequency and angular frequency bandwidth of the target bandpass filter based on the lower cutoff frequency and the upper cutoff frequency, and replacing the complex variables of the low-pass prototype with complex variables of the bandpass domain using complex frequency domain mapping rules.

[0009] Furthermore, converting the analog bandpass transfer function into a digital transfer function in the z-domain specifically involves using a bilinear transform to discretize the analog bandpass transfer function into a digital transfer function in the z-domain. Furthermore, the ultrasonic feature wave of the target to be extracted is a longitudinal wave signal.

[0010] Furthermore, the frequency band of the longitudinal wave signal is from 12MHz to 35MHz.

[0011] Furthermore, in a second aspect, the present invention provides a laser ultrasonic signal processing system for a dual-wave mixing interferometer, comprising: The acquisition module is used to acquire the laser ultrasonic time-domain signal received by the dual-wave mixing interferometer; The smoothing module is used to perform discrete wavelet decomposition on the laser ultrasound time-domain signal to obtain detail coefficients and approximation coefficients at each level; based on unbiased risk estimation, an adaptive threshold is determined for each level of detail coefficients, and the detail coefficients at each level are smoothed according to the soft threshold rule; The reconstruction module is used to perform inverse wavelet transform and reconstruction on the smoothed detail coefficients and the retained approximation coefficients to obtain the reconstructed signal; The extraction module is used to construct a digital bandpass filter with maximum flatness in the passband based on the spectral distribution of the target ultrasonic feature wave to be extracted; and to use the digital bandpass filter to extract the frequency band of the reconstructed signal to obtain the target ultrasonic feature signal.

[0012] A third aspect of the present invention provides an electronic device comprising: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the laser ultrasonic signal processing method for a dual-wave mixing interferometer provided in the first aspect of the present invention.

[0013] In a fourth aspect, the present invention provides a computer-readable medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the laser ultrasonic signal processing method for a two-wave mixing interferometer provided in the first aspect of the present invention.

[0014] The beneficial effects of this invention are: This invention represents a breakthrough in ensuring the fidelity of laser-ultrasonic signal characteristics. In quantitative analysis of nondestructive testing, the amplitude of the defect echo directly determines the accuracy of defect size and depth assessment. Existing technologies often employ Chebyshev or elliptic filters, whose inherent "ripples" within the passband severely distort the amplitude of weak signals. Based on the physical frequency band characteristics of laser-ultrasonic testing, this invention derives a customized Butterworth digital bandpass filtering algorithm. This algorithm exhibits excellent amplitude-frequency flatness characteristics within the target frequency band of 12-35MHz, ensuring that the high-frequency longitudinal waves carrying microscopic defect characteristics do not lose energy and have zero amplitude distortion during extraction, laying an extremely solid data foundation for subsequent high-precision and highly reliable quantitative defect inversion.

[0015] This method demonstrates excellent and stable wideband anti-interference capability when dealing with complex interferometer receiving environments. High-frequency longitudinal wave signals are easily attenuated and often hidden within extremely strong low-frequency signals. The energy of these low-frequency interferences (such as surface waves and transverse waves) and system high-frequency noise typically exhibits a monotonic variation with frequency. If conventional filters with insufficiently smooth stopband characteristics are used, local ripples can easily lead to the accidental amplification of interference signals at specific frequencies. The customized filtering model of this invention possesses a strictly monotonically increasing stopband attenuation characteristic. This characteristic can smoothly and stably suppress strong-energy low-frequency surface waves and transverse wave interference outside the passband, while simultaneously filtering out high-frequency white noise, completely eliminating the risk of abnormal amplification of noise outside the frequency band, and greatly improving the extraction purity of the useful longitudinal wave signal with a center frequency of approximately 16MHz.

[0016] Based on the aforementioned highly fidelity and interference-resistant signal processing framework, this method was validated in tests on a 1mm thick 6061 aluminum alloy standard sample. Experimental data, after denoising and filtering using the method of this invention, showed that the actual measurement time for the longitudinal wave to reach the non-defect point was approximately 0.1672μs. Compared to the theoretical propagation time of 0.1613μs for this material, the time error was only 3.66%. This extremely small calculation error strongly verifies the rationality and scientific validity of this signal processing method, and can directly and significantly improve the detection capability and quantification accuracy of the dual-wave mixer interferometer for minute defects within materials. Attached Figure Description

[0017] Figure 1 This is a basic flowchart illustrating the laser ultrasonic signal processing method for a dual-wave mixing interferometer in some embodiments of the present invention. Figure 2 This is a schematic diagram comparing laser ultrasound signals before and after wavelet transform denoising in some embodiments of the present invention; Figure 3 This is a schematic diagram showing the amplitude and phase response of a pass filter in some embodiments of the present invention; Figure 4This is a schematic diagram comparing the amplitude and phase responses of four filters in some embodiments of the present invention; Figure 5 This is a schematic diagram comparing the time-domain signals before and after the bandpass filter in some embodiments of the present invention, wherein (a) and (b) are the time-domain signals before and after the bandpass filter, respectively; Figure 6 This is a schematic diagram comparing the laser ultrasound spectrum analysis before and after bandpass filtering in some embodiments of the present invention, wherein (a) and (b) are the laser ultrasound spectrum signals before and after bandpass filtering, respectively; Figure 7 This is a schematic diagram of the structure of a laser ultrasonic signal processing system for a dual-wave mixing interferometer in some embodiments of the present invention; Figure 8 This is a schematic diagram of the structure of an electronic device in some embodiments of the present invention. Detailed Implementation

[0018] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0019] Example 1 refer to Figure 1 and Figure 2 In a first aspect of the present invention, a laser ultrasonic signal processing method for a two-wave mixing interferometer is provided, comprising: S100. Acquire the laser ultrasonic time-domain signal received by the dual-wave mixing interferometer; S200. Perform discrete wavelet decomposition on the laser ultrasound time-domain signal to obtain detail coefficients and approximation coefficients at each level; determine an adaptive threshold for each level of detail coefficients based on unbiased risk estimation, and smooth the detail coefficients at each level according to the soft threshold rule; S300. Perform inverse wavelet transform and reconstruction on the smoothed detail coefficients and the retained approximation coefficients to obtain the reconstructed signal; S400. Based on the spectral distribution of the target ultrasonic feature wave to be extracted, a digital bandpass filter with maximum flatness in the passband is constructed; the reconstructed signal is extracted using the digital bandpass filter to obtain the target ultrasonic feature signal.

[0020] In step S100 of some embodiments of the present invention, the laser ultrasonic time-domain signal received by the dual-wave mixing interferometer is acquired; Specifically, the original ultrasonic time-domain signal generated by laser excitation on the surface of the target sample is received using a dual-wave mixer interferometer. This original time-domain signal is mixed with high-intensity low-frequency interference (such as surface waves and transverse waves), high-frequency noise generated by the system, and easily attenuated useful high-frequency signals (such as longitudinal waves that reflect the microscopic defects inside the material).

[0021] It is understandable that, since laser ultrasonic signals have typical non-stationary and transient change characteristics in the time domain, if they are directly input into a conventional global frequency domain filter, the Gibbs effect is very likely to be triggered, causing the time domain peak (i.e., the arrival time reflecting the depth of the defect) to become blurred or broadened.

[0022] Therefore, before frequency domain bandpass filtering, this invention innovatively introduces a time-frequency localized amplitude-preserving denoising algorithm based on wavelet transform.

[0023] In step S200 of some embodiments of the present invention, the laser ultrasound time-domain signal is subjected to discrete wavelet decomposition to obtain detail coefficients and approximation coefficients of each layer; an adaptive threshold is determined for each layer detail coefficient based on unbiased risk estimation, and the detail coefficients of each layer are smoothed according to the soft threshold rule; The laser ultrasound time-domain signal is subjected to discrete wavelet decomposition to obtain the detail coefficients and approximation coefficients of each layer, including discrete wavelet decomposition and Symlets basis function mapping. Specifically, the noisy one-dimensional discrete ultrasonic signal acquired by the interferometer is set as The Mallat pyramid algorithm was used to perform multi-scale discrete wavelet transform (DWT) on the signal.

[0024] In selecting the wavelet basis, this embodiment preferably uses Symlets 4 (sym4) as the mother wavelet. Compared with the conventional Daubechies wavelet, the Symlets wavelet family maintains compact support characteristics while possessing better approximate symmetry. This symmetry can maximize the guarantee of strictly linear phase of the ultrasonic wave train during decomposition and reconstruction, preventing phase shift distortion of the time-domain signal. The decomposition level is set to 5 levels. In layer decomposition, the signal passes through a low-pass filter. and high-pass filter Producing approximate coefficients With detail coefficient The difference equation for its decomposition process is expressed as: , , in, This is the translation factor. Through a 5-layer decomposition, the high-frequency broadband white noise interwoven in the original signal can be completely stripped down to the detail coefficients of each layer. middle.

[0025] The step of determining the adaptive threshold corresponding to each level of detail coefficients based on unbiased risk estimation includes: constructing an unbiased risk estimation function corresponding to a candidate threshold for each level of detail coefficients; searching for a candidate threshold that minimizes the unbiased risk estimation function; and using the candidate threshold obtained from the search as the optimal denoising threshold for that level of detail coefficients.

[0026] Specifically, for broadband high-frequency noise with complex sources and unknown variance in laser ultrasound systems, traditional fixed threshold methods are prone to over-filtering or under-filtering. This embodiment employs the rigrsure algorithm model of Stein's Unbiased Risk Estimate (SURE) to evaluate the high-frequency detail coefficients at each layer. Perform global adaptive threshold calculation.

[0027] Let the detail coefficient vector of a certain layer be... For a given candidate threshold The risk estimation function defined by the SURE algorithm for: , In the formula, This represents the total number of detail coefficients in this layer. It is a counting term. It represents the coefficient whose absolute value is less than or equal to the threshold among all N coefficients. The number of coefficients (these coefficients are treated as noise and directly zeroed in soft thresholding). This is an energy / bias term. For each coefficient, take its absolute value. and threshold The smaller of the two values ​​is then squared and summed. This represents the amount of signal energy reduction after thresholding.

[0028] The Rigrsure algorithm dynamically searches for the optimal threshold that minimizes the mean squared error by minimizing the aforementioned risk estimation function. : , What we are looking for is the parameter that minimizes the risk. The algorithm itself can adaptively adjust the threshold according to the actual signal-to-noise ratio of the ultrasound signal, and has extremely strong robustness.

[0029] The smoothing process of the detail coefficients of each layer according to the soft threshold rule includes: setting the detail coefficients whose absolute value is less than or equal to the corresponding adaptive threshold to zero, and shrinking the detail coefficients whose absolute value is greater than the corresponding adaptive threshold to zero while keeping the sign unchanged.

[0030] Specifically, in obtaining the optimal threshold Then, soft thresholding (sorh = 's') is applied to the high-frequency detail coefficients of each layer. Unlike hard thresholding, which directly sets coefficients below the threshold to zero, the soft thresholding function smoothly shrinks the retained coefficients. Its mathematical expression is: , In the formula, The sign function is used. Soft thresholding allows wavelet coefficients to smoothly transition to zero, completely eliminating signal artifacts and additional oscillations caused by hard thresholding from the underlying mathematical logic.

[0031] In step S300 of some embodiments of the present invention, the smoothed detail coefficients and the retained approximation coefficients are subjected to inverse wavelet transform and reconstruction to obtain the reconstructed signal; Specifically, after processing the detail coefficients, the low-frequency approximation coefficients of layer 5 are forcibly retained (i.e., parameter keepapp = 1) to ensure that the low-frequency macroscopic morphology of the ultrasound signal is not lost. Finally, the processed detail coefficients are used... Approximation coefficients retained Reconstruction is performed using the one-dimensional discrete wavelet inverse transform (IDWT): , After the inverse transformation described above, the reconstructed ultrasonic time-domain signal is finally output. This signal effectively filters out broadband white noise, and the transient peak and amplitude characteristics of the ultrasonic longitudinal wave are perfectly preserved. It is then used as a high-fidelity "initial screening signal" to be input into a subsequent customized digital bandpass filter for further targeted frequency band extraction.

[0032] In step S400 of some embodiments of the present invention, a digital bandpass filter with maximum flatness in the passband is constructed based on the spectral distribution of the target ultrasonic feature wave to be extracted; the reconstructed signal is then subjected to frequency band extraction using the digital bandpass filter to obtain the target ultrasonic feature signal.

[0033] The construction of a digital bandpass filter with maximum flatness in the passband based on the spectral distribution of the ultrasonic feature wave of the target to be extracted includes: S401. Construct the prototype transfer function of Butterworth low-pass filter based on the cutoff angular frequency and the filter order; Specifically, to ensure that the ultrasonic echo amplitude, which reflects the size of the micropore defects, is not distorted, this invention constructs the square function of the amplitude-frequency response of the low-pass prototype filter using "maximum passband flatness" as a constraint: , In the formula, The amplitude response of the filter. The angular frequency of the signal. The cutoff angular frequency, The filter order is given. The mathematical characteristics of this prototype ensure extremely smooth and ripple-free signal transmission within the passband, avoiding the signal "distortion" caused by conventional Chebyshev filters.

[0034] S402. Based on the lower cutoff frequency and upper cutoff frequency of the ultrasonic feature wave of the target to be extracted, the Butterworth low-pass prototype transfer function is mapped to an analog bandpass transfer function; Furthermore, the step of mapping the Butterworth low-pass prototype transfer function to an analog bandpass transfer function includes: determining the center angular frequency and angular frequency bandwidth of the target bandpass filter based on the lower cutoff frequency and the upper cutoff frequency, and replacing the complex variables of the low-pass prototype with complex variables of the bandpass domain using complex frequency domain mapping rules.

[0035] Specifically, based on the spectral characteristics of the interferometer-received signal in this embodiment, the high-frequency longitudinal wave is mainly concentrated in the 12-35MHz frequency band. Therefore, this invention sets the lower cutoff frequency of the filter. Upper limit cutoff frequency .

[0036] The center angular frequency of the target bandpass filter can then be calculated. With angular frequency bandwidth : , , To achieve accurate extraction of this specific frequency band, this invention introduces a complex frequency domain mapping rule to transform the low-pass prototype transfer function. Complex variables in By making the following substitutions, a customized analog bandpass transfer function is derived. : , This mapping not only precisely shifts the flat passband to the 12-35 MHz range, but also ensures a strict monotonic attenuation characteristic for surface waves / transverse waves below 12 MHz and system noise above 35 MHz.

[0037] S403. The analog bandpass transfer function is converted into a digital transfer function in the z-domain using a bilinear transform to obtain the digital bandpass filter.

[0038] Specifically, considering that the signal acquired by the interferometer is a discrete digital signal, this embodiment further employs a bilinear transform to transform the aforementioned analog bandpass transfer function. Discretized digital transfer function in the z-domain : , In the formula, Let be the sampling period of the data acquisition system. After substituting and simplifying, we can obtain the digital filter transfer function format used for practical signal processing: , Taking into account computational delay, phase linearity, and stopband attenuation rate, this embodiment optimizes the order of the prototype filter to be [value missing]. (The corresponding bandpass digital filter is an 8th-order IIR filter). At this order, the corresponding filter coefficients are extracted. and Finally, a difference equation for point-by-point processing of time-domain signals is established: , In the formula, The input is the original signal sequence of the interferometer. The processed output signal sequence, M This indicates the order of the IIR digital filter.

[0039] Next, the reconstructed signal is subjected to frequency band extraction using the digital bandpass filter to obtain the target ultrasonic feature signal.

[0040] Specifically, the original laser ultrasound time-domain signal is input into the aforementioned digital difference equation, which is custom-designed according to this invention, for computational processing. The results show that this custom filter can correctly extract high-frequency longitudinal wave signals with a bandwidth of 12-35 MHz and a center frequency of approximately 16 MHz, completely eliminating low-frequency wave train interference and suppressing high-frequency noise.

[0041] In the verification experiment on 6061 aluminum alloy, the time for the longitudinal wave to reach the non-defect point, after filtering and extraction, was accurately measured to be 0.1672 μs. Figure 5 As shown, the error is only 3.66% compared to the theoretical propagation time of 0.1613 μs. This extremely high time resolution and amplitude reproduction fully verify the scientific nature and engineering practical value of this invention, which starts from the underlying physical characteristics and custom-derives the mathematical model of the filter.

[0042] It is understood that the primary key point of this invention lies in breaking away from the traditional thinking of single frequency domain filtering or time domain processing, and constructing a cascaded processing framework of "wavelet pre-screening + differential equation targeted extraction". Addressing the complex non-stationary characteristics of laser ultrasound signals, which combine transient changes with broadband noise masking, this architecture effectively avoids the time-domain peak blurring caused by direct global frequency domain filtering. While perfectly preserving the transient physical characteristics of weak high-frequency longitudinal waves, it achieves the step-by-step and complete removal of interference wave trains.

[0043] This invention delves into the underlying mathematical logic, abandoning empirically fixed thresholds that easily lead to signal overkill, and innovatively introduces a rigrsure dynamic optimization algorithm based on unbiased risk estimation (SURE). The key point is: within a selected approximately symmetric sym4 wavelet basis and a 5-level decomposition space, the risk estimation function is minimized. It automatically locks the optimal noise reduction threshold. Combined with soft threshold smoothing and shrinking processing, it fundamentally eliminates signal artifacts caused by hard thresholds, and achieves "zero damage" preservation of the peak amplitude of defect echoes while filtering out high-frequency broadband white noise.

[0044] Addressing the critical drawbacks of conventional Chebyshev or elliptic filters, such as "ripples" in the passband and the tendency to distort the amplitude of echoes from minute defects, this invention, starting from the frequency band characteristics of laser ultrasonic longitudinal waves, derives a customized maximum flatness digital bandpass algorithm. Its key breakthrough lies in: precisely targeting the 12-35 MHz frequency band; ensuring absolute flatness (zero distortion) of signal energy within the passband through complex frequency domain mapping and bilinear transformation from analog low-pass to digital bandpass; and utilizing its strictly monotonically increasing stopband characteristics to smoothly and stably suppress low-frequency surface wave and shear wave interference.

[0045] Example 2 refer to Figure 7 A second aspect of the present invention provides a laser ultrasonic signal processing system 1 for a two-wave mixing interferometer, comprising: Acquisition module 11 is used to acquire the laser ultrasonic time-domain signal received by the dual-wave mixing interferometer; The smoothing module 12 is used to perform discrete wavelet decomposition on the laser ultrasound time-domain signal to obtain detail coefficients and approximation coefficients at each level; to determine an adaptive threshold for each level of detail coefficients based on unbiased risk estimation; and to smooth the detail coefficients at each level according to the soft threshold rule. Reconstruction module 13 is used to perform inverse wavelet transform and reconstruction on the smoothed detail coefficients and the retained approximation coefficients to obtain the reconstructed signal; Extraction module 14 is used to construct a digital bandpass filter with maximum flatness in the passband based on the spectral distribution of the target ultrasonic feature wave to be extracted; and to use the digital bandpass filter to extract the frequency band of the reconstructed signal to obtain the target ultrasonic feature signal.

[0046] Furthermore, the smoothing module 12 includes: a construction unit for constructing an unbiased risk estimation function corresponding to a candidate threshold for each level of detail coefficients; a search unit for searching for a candidate threshold that minimizes the unbiased risk estimation function; and using the searched candidate threshold as the optimal denoising threshold for that level of detail coefficients.

[0047] Example 3 refer to Figure 8 A third aspect of the present invention provides an electronic device comprising: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the laser ultrasonic signal processing method for a two-wave mixing interferometer of the present invention in the first aspect.

[0048] Electronic device 500 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 502 or a program loaded from storage device 508 into random access memory (RAM) 503. The RAM 503 also stores various programs and data required for the operation of electronic device 500. The processing unit 501, ROM 502, and RAM 503 are interconnected via bus 504. An input / output (I / O) interface 505 is also connected to bus 504.

[0049] Typically, the following devices can be connected to I / O interface 505: input devices 506 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 507 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 508 including, for example, hard disks; and communication devices 509. Communication device 509 allows electronic device 500 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 8 An electronic device 500 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 8 Each box shown can represent a device or multiple devices as needed.

[0050] Specifically, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 509, or installed from a storage device 508, or installed from a ROM 502. When the computer program is executed by a processing device 501, it performs the functions defined in the methods of embodiments of this disclosure. It should be noted that the computer-readable medium described in embodiments of this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In embodiments of this disclosure, a computer-readable storage medium may 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 embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may 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. Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0051] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more computer programs, which, when executed by the electronic device, cause the electronic device to: Computer program code for performing the operations of embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, C++, and Python—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0052] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0053] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A laser ultrasonic signal processing method for a two-wave mixing interferometer, characterized in that, include: Acquire the laser ultrasonic time-domain signal received by the dual-wave mixing interferometer; Discrete wavelet decomposition is performed on the laser ultrasound time-domain signal to obtain detail coefficients and approximation coefficients at each level; an adaptive threshold is determined for each level of detail coefficients based on unbiased risk estimation, and the detail coefficients at each level are smoothed according to the soft threshold rule; The smoothed detail coefficients and the retained approximation coefficients are subjected to inverse wavelet transform and reconstruction to obtain the reconstructed signal; Based on the spectral distribution of the target ultrasonic feature wave to be extracted, a digital bandpass filter with maximum flatness in the passband is constructed; the reconstructed signal is then subjected to frequency band extraction using the digital bandpass filter to obtain the target ultrasonic feature signal.

2. The laser ultrasonic signal processing method for a dual-wave mixing interferometer according to claim 1, characterized in that, The method for determining the adaptive thresholds corresponding to the detail coefficients of each layer based on unbiased risk estimation includes: For each level of detail coefficients, construct an unbiased risk estimation function corresponding to the candidate threshold; Search for candidate thresholds that minimize the unbiased risk estimation function; The candidate threshold obtained from the search is used as the optimal denoising threshold for the detail coefficient of that layer.

3. The laser ultrasonic signal processing method for a dual-wave mixing interferometer according to claim 1, characterized in that, The smoothing process of the detail coefficients of each layer according to the soft threshold rule includes: Detail coefficients whose absolute value is less than or equal to the corresponding adaptive threshold are set to zero, while detail coefficients whose absolute value is greater than the corresponding adaptive threshold are shrunk to zero and their signs remain unchanged.

4. The laser ultrasonic signal processing method for a dual-wave mixing interferometer according to claim 1, characterized in that, The construction of a digital bandpass filter with maximum flatness in the passband based on the spectral distribution of the ultrasonic feature wave of the target to be extracted includes: Construct the prototype transfer function of Butterworth low-pass filter based on the cutoff angular frequency and the filter order; Based on the lower and upper cutoff frequencies of the ultrasonic feature wave of the target to be extracted, the Butterworth low-pass prototype transfer function is mapped to an analog bandpass transfer function. The analog bandpass transfer function is converted into a digital transfer function in the z-domain to obtain the digital bandpass filter.

5. The laser ultrasonic signal processing method for a dual-wave mixing interferometer according to claim 4, characterized in that, The step of mapping the Butterworth low-pass prototype transfer function to an analog bandpass transfer function includes: determining the center angular frequency and angular frequency bandwidth of the target bandpass filter based on the lower cutoff frequency and the upper cutoff frequency, and replacing the complex variables of the low-pass prototype with complex variables of the bandpass domain using complex frequency domain mapping rules.

6. The laser ultrasonic signal processing method for a dual-wave mixing interferometer according to claim 4, characterized in that, Specifically, converting the analog bandpass transfer function into a digital transfer function in the z-domain involves using a bilinear transformation to discretize the analog bandpass transfer function into a digital transfer function in the z-domain.

7. The laser ultrasonic signal processing method for a dual-wave mixing interferometer according to claim 4, characterized in that, The ultrasonic feature wave of the target to be extracted is a longitudinal wave signal.

8. The laser ultrasonic signal processing method for a dual-wave mixing interferometer according to claim 7, characterized in that, The frequency band of the longitudinal wave signal is from 12MHz to 35MHz.

9. The laser ultrasonic signal processing method for a dual-wave mixing interferometer according to claim 1, characterized in that, The laser ultrasonic signal processing system described below is used, and the system includes: The acquisition module is used to acquire the laser ultrasonic time-domain signal received by the dual-wave mixing interferometer; The smoothing module is used to perform discrete wavelet decomposition on the laser ultrasound time-domain signal to obtain detail coefficients and approximation coefficients at each level; based on unbiased risk estimation, an adaptive threshold is determined for each level of detail coefficients, and the detail coefficients at each level are smoothed according to the soft threshold rule; The reconstruction module is used to perform inverse wavelet transform and reconstruction on the smoothed detail coefficients and the retained approximation coefficients to obtain the reconstructed signal; The extraction module is used to construct a digital bandpass filter with maximum flatness in the passband based on the spectral distribution of the target ultrasonic feature wave to be extracted; and to use the digital bandpass filter to extract the frequency band of the reconstructed signal to obtain the target ultrasonic feature signal.

10. The laser ultrasonic signal processing method for a dual-wave mixing interferometer according to claim 9, characterized in that, The smoothing module includes: The building unit is used to construct an unbiased risk estimation function corresponding to the candidate threshold for each level of detail coefficients; The search unit is used to search for candidate thresholds that minimize the unbiased risk estimation function; the candidate thresholds obtained by the search are used as the optimal denoising thresholds for the detail coefficients of that layer.

Images