A seismic signal denoising method, device and storage medium

Abstract

The application discloses a seismic signal denoising method and device and a storage medium, and relates to the technical field of signal processing.The technical scheme points of the method are as follows: the method comprises the following steps: acquiring an original collected seismic signal; performing wavelet transform threshold denoising on the seismic signal to obtain a first signal sequence from which white noise is removed; detecting the first signal sequence to determine the positions of each pulse noise in the first signal sequence and marking the positions as pulse noise points; and performing median filtering processing on each pulse noise point in the first signal sequence to obtain a second signal sequence from which pulse noise is removed as a denoised seismic signal sequence.Through the denoising method, white noise can be efficiently removed, pulse noise can be removed in a targeted manner, the original information of the signal can be retained to the maximum extent, and the preprocessing quality of the seismic signal is improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and specifically to a method, apparatus and storage medium for denoising seismic signals. Background Technology

[0002] This section is intended to provide background or context for the embodiments set forth in the claims. The description herein is not an admission that it is prior art simply because it is included in this section.

[0003] In signal acquisition, such as during seismic acquisition, the raw seismic data acquired by seismic nodal instruments is often mixed with irregular noise such as white noise and impulse noise due to the influence of natural and human environments, coupled with the weak instantaneous excitation energy of controllable seismic sources. To remove the influence of white noise and impulse noise, traditional methods typically employ median filtering algorithms. This mainly involves applying median filtering to each raw signal sample in the time or frequency domain, effectively removing white noise and impulse noise. However, simply using median filtering can also lead to over-denoising, resulting in many identical signal levels and severe signal distortion. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a method, apparatus, and storage medium for denoising seismic signals. This method effectively removes white noise and specifically removes impulse noise. Compared with traditional median filtering algorithms that remove both white noise and impulse noise from the original signal simultaneously, this method retains more original signal information and improves the preprocessing quality of seismic signals.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention includes four aspects.

[0006] Firstly, a method for denoising seismic signals is provided, comprising the following steps:

[0007] Acquire the raw seismic signals;

[0008] The seismic signal is subjected to wavelet transform threshold denoising to obtain a first signal sequence with white noise removed;

[0009] The first signal sequence is detected to determine the position of each impulse noise in the first signal sequence and marked as an impulse noise point;

[0010] Median filtering is performed on each impulse noise point in the first signal sequence to obtain a second signal sequence with impulse noise removed, which is then used as the denoised seismic signal sequence.

[0011] In some embodiments, performing wavelet transform threshold denoising on the seismic signal to obtain a first signal sequence with white noise removed includes:

[0012] The seismic signal is subjected to time-frequency transformation using Morlet wavelet transform to obtain the time spectrum of the seismic signal;

[0013] The noise reduction threshold is adaptively determined based on the characteristics of the seismic signal in the time-frequency domain.

[0014] White noise is removed from the time spectrum according to the denoising threshold to obtain the wavelet coefficient matrix after thresholding.

[0015] The wavelet coefficient matrix is ​​reconstructed by inverse wavelet transform to obtain the first signal sequence after removing white noise.

[0016] In some embodiments, the step of adaptively determining the denoising threshold based on the characteristics of the seismic signal in the time-frequency domain includes:

[0017] The noise standard deviation is adaptively obtained based on the characteristics of the seismic signal in the time-frequency domain.

[0018] The noise reduction threshold is determined based on the noise standard deviation.

[0019] The noise standard deviation is denoted as σ, and the formula for the noise standard deviation is: σ = M / 0.6745;

[0020] In the formula, M is the median of the wavelet transform coefficient modulus of the signal at the minimum scale;

[0021] The denoising threshold is denoted as T, and the formula for the denoising threshold is: T=σ(√2ln(n));

[0022] In the formula, ln(n) represents taking the natural logarithm with respect to the signal length n.

[0023] In some embodiments, the step of removing white noise from the time spectrum according to the denoising threshold to obtain a thresholded wavelet coefficient matrix includes:

[0024] The wavelet coefficient matrix to be processed is determined based on the time spectrum.

[0025] Based on the scale factor and time factor, the wavelet coefficient matrix to be processed is traversed, and the wavelet coefficients with a modulus value less than the denoising threshold are determined as the wavelet coefficients corresponding to white noise.

[0026] By setting the wavelet coefficients corresponding to the white noise to zero, the wavelet coefficient matrix after thresholding is obtained.

[0027] In some embodiments, detecting the first signal sequence, determining the position of each impulse noise in the first signal sequence, and marking it as an impulse noise point includes:

[0028] Determine the length of the detection window;

[0029] Several detection points are set according to the first signal sequence;

[0030] Based on the detection window length, each detection point is analyzed and processed sequentially, and the detection points that satisfy the preset discriminant are determined as impulse noise location points;

[0031] Each impulse noise location point is marked as an impulse noise point.

[0032] In some embodiments, the step of sequentially analyzing and processing each detection point according to the detection window length, and determining the detection point that satisfies the preset discriminant as the impulse noise location point, includes:

[0033] Perform the following sliding process on each point to be detected sequentially until all impulse noise points are marked:

[0034] Based on the detection window length and the position of the point to be detected, determine several neighboring signal sample points of the point to be detected within the detection window length range, and aggregate them into neighboring points;

[0035] The differences between the point to be detected and each neighboring signal sample point are analyzed and processed to determine the number of differences that are greater than a preset difference threshold.

[0036] Determine whether the ratio of the number of differences to the number of neighbor signal samples is greater than a preset percentage threshold;

[0037] If so, mark the point to be detected as a pulse noise location point.

[0038] In some embodiments, the discriminant for determining the location of the impulse noise is as follows:

[0039]

[0040] In the formula, y(t) i ) represents the point to be detected; y(t) represents the neighboring signal sample points; N(t) represents the signal sample points. i ) represents the neighboring points of the point to be detected within the window length range; Tv is the threshold for judging the difference between the point to be detected and the neighboring signal samples; Tn is the threshold for judging the number of large differences between the point to be detected and the neighboring samples.

[0041] In some embodiments, performing median filtering on the impulse noise points in the first signal sequence to obtain a second signal sequence with impulse noise removed includes:

[0042] Using the location of the impulse noise point as the center, extract the first segment of the signal with a window length of win;

[0043] Sort the first segment of signal, and take the median and the values ​​that are adjacent to the median and are less than the median, and sum and average them to obtain the two-point mean;

[0044] Replace the impulse noise points in the original acquired signal with the average of two points;

[0045] The above steps are repeated for each impulse noise point in the first signal sequence, and a sliding window process is performed to obtain a second signal sequence with impulse noise removed.

[0046] Secondly, a seismic signal denoising device is provided, comprising:

[0047] The signal acquisition module is used to acquire the raw seismic signals.

[0048] The white noise processing module is used to perform wavelet transform threshold denoising on the seismic signal to obtain a first signal sequence with white noise removed.

[0049] The signal processing module is used to detect the first signal sequence, determine the position of each impulse noise in the first signal sequence, and mark it as an impulse noise point;

[0050] The impulse noise processing module is used to perform median filtering on each impulse noise point in the first signal sequence to obtain a second signal sequence with impulse noise removed, which is then used as the denoised seismic signal sequence.

[0051] Thirdly, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the aforementioned noise reduction method.

[0052] Fourthly, a computer-readable storage medium is provided, wherein when the computer program is executed by a processor, it implements the steps of the aforementioned denoising method.

[0053] Compared with the prior art, one or more embodiments of the above solutions may have the following advantages or beneficial effects:

[0054] This application provides a method, apparatus, and storage medium for denoising seismic signals. The denoising method includes the following steps: acquiring the original acquired seismic signal; performing wavelet transform threshold denoising on the seismic signal to obtain a first signal sequence with white noise removed; detecting the first signal sequence to determine the position of each impulse noise point within the first signal sequence and marking them as impulse noise points; and performing median filtering on each impulse noise point in the first signal sequence to obtain a second signal sequence with impulse noise removed, which is then used as the denoised seismic signal sequence. This denoising method effectively removes white noise using wavelet transform threshold denoising technology, and further improves the removal of impulse noise by applying median filtering based on the position of the impulse noise points. It not only efficiently removes white noise but also specifically removes impulse noise. Compared to traditional median filtering algorithms that simultaneously remove both white noise and impulse noise from the original signal, this method retains more original signal information and improves the preprocessing quality of seismic signals. Attached Figure Description

[0055] The present application will be described in more detail below based on embodiments and with reference to the accompanying drawings;

[0056] Figure 1 This is a waveform diagram illustrating the result of median filtering for removing white noise and impulse noise in existing technologies.

[0057] Figure 2 This is an exemplary flowchart of the seismic signal denoising method in Embodiment 1 of the present invention;

[0058] Figure 3 In the embodiments of the present invention, corresponding to Figure 2 An exemplary flowchart of step S2 shown;

[0059] Figure 4 In the embodiments of the present invention, corresponding to Figure 3 An exemplary flowchart of step S22 shown;

[0060] Figure 5 In the embodiments of the present invention, corresponding to Figure 3 An exemplary flowchart of step S23 shown;

[0061] Figure 6 In the embodiments of the present invention, corresponding to Figure 2 An exemplary flowchart of step S3 shown;

[0062] Figure 7 In the embodiments of the present invention, corresponding to Figure 6 An exemplary flowchart of step S33 shown;

[0063] Figure 8 In the embodiments of the present invention, corresponding to Figure 2An exemplary flowchart of step S4 shown;

[0064] Figure 9 This is a waveform diagram illustrating the results of removing white noise and impulse noise using the denoising method according to an embodiment of the present invention.

[0065] Figure 10 This is a schematic block diagram of a seismic signal denoising device provided in Embodiment 2 of the present invention;

[0066] Figure 11 This is a schematic block diagram of an electronic device provided in Embodiment 3 of the present invention;

[0067] Figure 12 This is a schematic diagram of a computer-readable storage medium provided in Embodiment 4 of the present invention.

[0068] In the accompanying drawings, the same parts are referred to by the same reference numerals, and the drawings are not drawn to scale. Detailed Implementation

[0069] The present disclosure will be further described below with reference to the embodiments shown in the accompanying drawings.

[0070] Example 1:

[0071] In the field of signal acquisition, the acquired raw signals are often mixed with irregular noise such as white noise and impulse noise. To remove the influence of white noise and impulse noise, existing methods usually employ median filtering algorithms, applying median filtering to each sample point of the raw signal in the time or frequency domain to remove both white noise and impulse noise simultaneously. However, simply applying median filtering can lead to over-denoising, resulting in many signals with the same level or value. Over-denoising severely distorts the signal, such as... Figure 1 As shown, the median filtering in the prior art removes too much of the original signal when removing white noise and impulse noise, thus affecting signal quality.

[0072] This application discloses a method for denoising seismic signals, such as... Figure 2As shown, the method includes the following steps: acquiring the original acquired seismic signal; performing wavelet transform threshold denoising on the seismic signal to obtain a first signal sequence with white noise removed; detecting the first signal sequence to determine the position of each impulse noise point in the first signal sequence and marking them as impulse noise points; and performing median filtering on each impulse noise point in the first signal sequence to obtain a second signal sequence with impulse noise removed, which is then used as the denoised seismic signal sequence. This method effectively removes white noise using wavelet transform threshold denoising technology, and further improves the removal of impulse noise by performing median filtering based on the position of the impulse noise points. It not only efficiently removes white noise but also specifically removes impulse noise. Compared with traditional median filtering algorithms that simultaneously remove both white noise and impulse noise from the original signal, this method retains more original signal information and improves the preprocessing quality of seismic signals.

[0073] This disclosure provides at least one embodiment of a seismic signal denoising method, which is not limited to denoising seismic signals. This method can also be used to denoise other signals requiring white noise and impulse noise removal. The seismic signal denoising method can be implemented in software, hardware, firmware, or any combination thereof. It is loaded and executed by a processor in devices such as mobile phones, tablets, laptops, desktop computers, and web servers, thereby achieving efficient removal of white noise and targeted removal of impulse noise, retaining more original signal information, and improving the preprocessing quality of seismic signals.

[0074] The following is for reference. Figure 2 As shown, a seismic signal denoising method provided in at least one embodiment of the present disclosure will be described, the denoising method including steps S1 to S4.

[0075] S1. Obtain the original seismic signals.

[0076] In some embodiments, a seismic nodal instrument can be used to collect seismic signals in the target area, which are then transmitted over a network to the implementation carrier of the denoising method for analysis and processing.

[0077] S2. Perform wavelet transform threshold denoising on the seismic signal to obtain a first signal sequence with white noise removed.

[0078] In some embodiments, step S2, such as Figure 3 As shown, steps S21 to S24 are included:

[0079] S21. The seismic signal is subjected to time-frequency transformation using Morlet wavelet transform to obtain the time spectrum of the seismic signal.

[0080] In some embodiments, the complex Morlet wavelet basis is used as the wavelet basis function, and the wavelet basis function is denoted as:

[0081]

[0082] In the formula, I is the false alarm unit, ω0 is the wavelet center angular frequency, and t is the sampling time;

[0083] The scale factor of the wavelet transform is a i =2 i / Nv The corresponding sequence is denoted as a = {a i}, i=1,2,...,J,J=L*Nv, where Nv is the musical note, which can take the value 32, and n is the signal length. Therefore, the wavelet transform expression in the time domain is:

[0084]

[0085] In the formula, x(t) represents the seismic signal sequence; "—" indicates that the function is taken as its conjugate; a i b is the scaling factor; j t is the time shift factor; t is the sampling time.

[0086] Furthermore, W(a) can be... i ,b j (abbreviated as W) i,j , representing the wavelet coefficients at the j-th time point under the i-th scale factor, to facilitate subsequent white noise processing.

[0087] Based on the above expressions, the calculation principle of wavelet transform in the time domain is the convolution of the signal with the wavelet function at each scale factor. Therefore, the calculation principle in the frequency domain is the direct conjugate product of the Fourier transform result of the signal and the frequency domain expression of the wavelet transform with scale factors. Performing the transformation in the frequency domain is more efficient than performing convolution calculations in the time domain. The specific operation is as follows:

[0088] Let X(ω) be the Fourier transform of signal x(t), and the specific formula is as follows:

[0089] X(ω)=fx(t)e -Iωt dt;

[0090] In the formula, ω represents the angular frequency, I is the imaginary unit, and t is the sampling time;

[0091] The inner product of functions f(τ) and g(τ) is:

[0092]

[0093] In the formula, "<>" represents the inner product of two functions, and "—" represents the functions taking their conjugates;

[0094] The convolution of functions f(τ) and g(τ) is:

[0095] f(t)*g(t)=ff(t)g(tt)dt;

[0096] In the formula, "*" represents the convolution symbol of the function;

[0097] Based on the definitions of inner product and convolution of the aforementioned functions, the wavelet transform in the time domain can be equivalent to the following second formula:

[0098]

[0099] The wavelet basis function is denoted as The Fourier transform is Based on the theory of function inversion and conjugation in Fourier transform:

[0100]

[0101] The Fourier transform is:

[0102]

[0103] This formula calculates the discrete values ​​of ω with respect to each scale factor in the range [0, 2π], yielding a wavelet function matrix A(i,j) with respect to the scale sequence a. According to Fourier transform theory, the time-domain convolution of two functions is equal to the frequency-domain product of their respective functions. Therefore, the calculation of the expression on the left side of the second formula, which is equivalent to the expression in the frequency domain, is equivalent to performing a dot product of each row of matrix A(i,j) with the matrix X(ω), thus obtaining the wavelet coefficient matrix W(a). i ,b j This process generates a time spectrum of the seismic signal, which facilitates white noise removal using the wavelet coefficient matrix and the time spectrum, thereby improving the efficiency of white noise removal alone.

[0104] S22. Adaptively determine the denoising threshold based on the characteristics of the seismic signal in the time and frequency domain.

[0105] In some embodiments, step S22, as follows Figure 4 As shown, it includes:

[0106] S221. Adaptively obtain the noise standard deviation based on the characteristics of the seismic signal in the time-frequency domain;

[0107] S222. Determine the denoising threshold based on the noise standard deviation.

[0108] Wherein, the noise standard deviation is denoted as σ, and the formula for the noise standard deviation is: σ=M / 0.6745;

[0109] In the formula, M is the median of the wavelet transform coefficient modulus of the signal at the minimum scale;

[0110] Let the denoising threshold be T, then the formula for the denoising threshold is:

[0111] In the formula, ln(n) represents taking the natural logarithm with respect to the signal length n.

[0112] In some embodiments, the denoising threshold T is obtained by using the median value of the wavelet coefficient modulus corresponding to the minimum scale, based on the wavelet coefficient matrix.

[0113] S23. White noise is removed from the time spectrum according to the denoising threshold to obtain the wavelet coefficient matrix after threshold processing.

[0114] In some embodiments, step S23, such as Figure 5 As shown, it includes:

[0115] S231. Determine the wavelet coefficient matrix to be processed based on the time spectrum;

[0116] S232. Based on the scale factor and time factor, traverse the wavelet coefficient matrix to be processed, and determine the wavelet coefficients with modulus values ​​less than the denoising threshold as the wavelet coefficients corresponding to white noise.

[0117] S233. Set the wavelet coefficients corresponding to the white noise to zero to obtain the wavelet coefficient matrix after thresholding.

[0118] S24. Perform inverse wavelet transform on the wavelet coefficient matrix to reconstruct the first signal sequence after removing white noise.

[0119] In some embodiments, the first signal sequence is denoted as y(i). The formula for reconstructing the first signal sequence after removing white noise by performing inverse wavelet transform on the wavelet coefficient matrix is ​​as follows:

[0120]

[0121] In the formula, the function real() represents taking the real part of the complex number; n is the signal length, i represents the index value corresponding to the i-th scale; j represents the j-th time shift factor; W ij This represents the wavelet coefficients at the j-th time point under the i-th scaling factor, which is also the element in the i-th row and j-th column of matrix W; A ij This represents the element in the i-th row and j-th column of matrix A; a i Let i be the i-th scaling factor.

[0122] S3. Detect the first signal sequence, determine the position of each impulse noise in the first signal sequence, and mark it as an impulse noise point.

[0123] In some embodiments, the salt-and-pepper noise detection method in the image field is used to detect impulse noise in the first signal sequence, which can accurately mark each impulse noise point in the first signal sequence, so as to accurately remove each impulse noise while preserving the original signal information and improve the signal processing quality.

[0124] In some embodiments, step S3, such as Figure 6 As shown, steps S31-S33 are included:

[0125] S31. Determine the length of the detection window.

[0126] In some embodiments, the detection window length is denoted as win, and is generally selected as 5, 7, 9, etc. In this embodiment, the detection window length range is set to 7. The first signal sequence is divided by the detection window length so that each impulse noise point can be marked by sliding window processing.

[0127] S32. Set several detection points according to the first signal sequence.

[0128] S33. Analyze and process each detection point in sequence according to the detection window length, and determine the detection points that meet the preset discrimination formula as impulse noise location points and mark them as impulse noise points.

[0129] In some embodiments, step S33, such as Figure 7 As shown, this includes: performing the following sliding process on each point to be detected sequentially until all impulse noise points are marked:

[0130] S331. Based on the detection window length and the position of the point to be detected, determine several neighboring signal sample points of the point to be detected within the detection window length range, and set them as neighbor points;

[0131] S332. Analyze and process the difference between the point to be detected and each neighboring signal sample point to determine the number of differences that are greater than the preset difference threshold.

[0132] S333. Determine whether the ratio of the number of differences to the number of neighbor signal samples is greater than a preset percentage threshold.

[0133] S334. If so, mark the point to be detected as a pulse noise location point.

[0134] In some embodiments, a preset difference threshold is denoted as Tv. In this embodiment, the threshold Tv = 30, used to determine whether the difference between the point to be detected and its neighboring signal samples is sufficiently large. A percentage threshold is denoted as Tn. In this embodiment... It is used to determine whether there are enough large differences between the point to be detected and its neighboring sample points.

[0135] In some embodiments, the discriminant for determining the location of the impulse noise is as follows:

[0136]

[0137] In the formula, y(t) i ) represents the point to be detected; y(t) represents the neighboring signal sample points; N(t) represents the signal sample points. i ) represents the neighboring points of the point to be detected within the window length range; Tv represents the threshold for judging the difference between the point to be detected and the neighboring signal samples; Tn represents the threshold for judging the number of large differences between the point to be detected and the neighboring samples; #[] represents the number of samples that satisfy [].

[0138] The discriminant can quickly identify impulse noise points within a set window length. By analyzing and processing each divided detection window length through sliding window processing, the location of impulse noise points in the first signal sequence can be quickly and accurately detected. This makes it easy to perform median filtering to remove impulse noise points only, thus preserving the original information of the signal to the greatest extent.

[0139] S4. Perform median filtering on each impulse noise point in the first signal sequence to obtain a second signal sequence with impulse noise removed, which is then used as the denoised seismic signal sequence.

[0140] In some embodiments, step S4, such as Figure 8 As shown, it includes:

[0141] S41. Using the location of the impulse noise point as the center, extract the first segment of the signal with a window length of win;

[0142] S42. Sort the first segment of signal, and take out the median and the value that is adjacent to the median and is less than the median, and sum and average them to obtain the mean of the two points;

[0143] S43. Replace the impulse noise points in the original acquired signal with the average of two points;

[0144] S44. Repeat the above steps for each impulse noise point in the first signal sequence and perform sliding window processing to obtain the second signal sequence with impulse noise removed.

[0145] In some embodiments, the detected impulse noise points are denoted as y(i), and the sequence...

[0146] Sort {y(i-3),...,y(i),...y(i+3)} in ascending order to obtain a new sequence {P1,P2,P3,P4,P5,P6,P7}. Then y(i) is replaced with (P3+P4) / 2. Mean averaging the median ensures the smoothness of the processed signal, thereby improving the preprocessing quality of seismic signals by preserving the original information of the signal to the greatest extent after impulse noise removal and ensuring the smoothness of the processed signal.

[0147] like Figure 9 As shown, the results of removing white noise and impulse noise using the seismic signal denoising method provided above demonstrate that the original information of the signal is preserved to the greatest extent, and the smoothness of the processed signal is guaranteed, thereby improving the preprocessing quality of the seismic signal.

[0148] The seismic signal denoising method provided in the embodiments of this disclosure can remove high-frequency white noise using wavelet transform while temporarily retaining impulse noise. It then further monitors the impulse noise in the signal sequence after removing white noise, accurately marking the position of each impulse noise point in the signal sequence. Finally, it removes each impulse noise point through median filtering, accurately and specifically removing impulse noise. This not only efficiently removes white noise but also specifically removes impulse noise, maximizing the retention of more original signal information and improving the preprocessing quality of seismic signal denoising, thus providing high-quality data for subsequent high-precision seismic data processing.

[0149] Example 2:

[0150] At least one embodiment of this disclosure also provides a seismic signal denoising device, the modules of which correspond to the steps in the above-described method. The effects achieved and the problems solved have been described, and will not be repeated here. This seismic signal denoising device, such as… Figure 10 As shown, it includes:

[0151] Signal acquisition module 11 is used to acquire the raw seismic signals;

[0152] White noise processing module 12 is used to perform wavelet transform threshold denoising on the seismic signal to obtain a first signal sequence with white noise removed.

[0153] The signal processing module 13 is used to detect the first signal sequence, determine the position of each impulse noise in the first signal sequence, and mark it as an impulse noise point;

[0154] The impulse noise processing module 14 is used to perform median filtering on each impulse noise point in the first signal sequence to obtain a second signal sequence with impulse noise removed as the denoised seismic signal sequence.

[0155] In some embodiments, the output terminal of the signal acquisition module 11 is connected to the input terminal of the white noise processing module 12, the output terminal of the white noise processing module 12 is connected to the input terminal of the signal processing module 13, and the output terminal of the signal processing module 13 is connected to the input terminal of the impulse noise processing module 14.

[0156] The output of the impulse noise processing module 14 can be connected to a display device (such as a monitor) to present the noise reduction results to the staff through the screen.

[0157] In some embodiments, the device may be part of software or implemented in conjunction with appropriate hardware, which will not be elaborated here.

[0158] Example 3:

[0159] At least some embodiments of this disclosure also provide an electronic device, such as Figure 11 As shown, the electronic device includes a memory 21 and a processor 22. The memory 21 stores a computer program that, when executed by the processor, performs the steps of a noise reduction method as provided in any embodiment of this disclosure.

[0160] In some embodiments, processor 22 is used to perform all or part of the steps in the denoising method as described in any embodiment of this disclosure. Memory 21 is used to store various types of data, which may include, for example, instructions for any application or method in an electronic device, as well as application-related data.

[0161] The processor 22 may be implemented as an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA), a controller, a microcontroller, a microprocessor, or other electronic components, and is used to execute the noise reduction method in Embodiment 1 above.

[0162] The memory 21 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0163] Example 4:

[0164] At least some embodiments of this disclosure also provide a computer-readable storage medium, such as Figure 12 As shown, the readable storage medium stores a computer program 31, which, when executed by a processor, implements the steps of the denoising method provided in any embodiment of this disclosure.

[0165] In some embodiments, the storage medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media may include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. More specific examples of machine-readable storage media include electrical connections based on 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.

[0166] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0167] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0168] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0169] Example 5:

[0170] This invention also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the denoising method provided in any embodiment of this disclosure.

[0171] In summary, the seismic signal denoising method, apparatus, and storage medium provided in this application can remove high-frequency white noise using wavelet transform while temporarily retaining impulse noise. Furthermore, the signal sequence after removing white noise is subjected to impulse noise monitoring to accurately mark the position of each impulse noise point in the signal sequence. Then, each impulse noise point is removed through median filtering. This accurate and targeted removal of impulse noise not only efficiently removes white noise but also specifically removes impulse noise, maximizing the retention of more original signal information and improving the preprocessing quality of seismic signal denoising. This provides high-quality data for subsequent high-precision seismic data processing.

[0172] The various embodiments in this disclosure are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0173] The scope of protection of this disclosure is not limited to the embodiments described above. Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from its scope and spirit. If such modifications and variations fall within the scope of the claims of this disclosure and their equivalents, then the intent of this disclosure also includes such modifications and variations.

Claims

1. A method for denoising seismic signals, characterized in that, Includes the following steps: Acquire the raw seismic signals; The seismic signal is subjected to wavelet transform threshold denoising to obtain a first signal sequence with white noise removed; The first signal sequence is detected to determine the position of each impulse noise in the first signal sequence and marked as an impulse noise point; The median filtering process is performed on each impulse noise point in the first signal sequence to obtain the second signal sequence after removing the impulse noise, which is then used as the denoised seismic signal sequence. The step of detecting the first signal sequence, determining the position of each impulse noise in the first signal sequence, and marking it as an impulse noise point includes: determining the detection window length; setting a number of detection points according to the first signal sequence; analyzing and processing each detection point according to the detection window length, and determining the detection points that satisfy the preset discriminant as impulse noise location points; and marking each impulse noise location point as an impulse noise point. The step of sequentially analyzing and processing each detection point according to the detection window length, and determining the detection points that satisfy the preset discriminant as impulse noise location points, includes: Perform the following sliding process on each point to be detected sequentially until all impulse noise points are marked: Based on the detection window length and the position of the point to be detected, determine several neighboring signal sample points of the point to be detected within the detection window length range, and aggregate them into neighboring points; The differences between the point to be detected and each neighboring signal sample point are analyzed and processed to determine the number of differences that are greater than a preset difference threshold. Determine whether the ratio of the number of differences to the number of neighbor signal samples is greater than a preset percentage threshold; If so, mark the point to be detected as a pulse noise location point. The discriminant for determining the location of the impulse noise is as follows: In the formula, Points to be tested; For neighbor signal samples; These are the neighboring points of the point to be detected within the window length; A preset difference threshold is used to determine whether the difference between the point to be detected and its neighboring signal samples is large enough; The percentage threshold for determining the number of points with large differences between the point to be detected and its neighboring points; #[ ] represents the number of points that satisfy [ ].

2. The seismic signal denoising method according to claim 1, characterized in that, The step of performing wavelet transform threshold denoising on the seismic signal to obtain a first signal sequence with white noise removed includes: The seismic signal is subjected to time-frequency transformation using Morlet wavelet transform to obtain the time spectrum of the seismic signal; The noise reduction threshold is adaptively determined based on the characteristics of the seismic signal in the time-frequency domain. White noise is removed from the time spectrum according to the denoising threshold to obtain the wavelet coefficient matrix after thresholding. The wavelet coefficient matrix is ​​reconstructed by inverse wavelet transform to obtain the first signal sequence after removing white noise.

3. The seismic signal denoising method according to claim 2, characterized in that, The adaptive determination of the denoising threshold based on the characteristics of the seismic signal in the time-frequency domain includes: The noise standard deviation is adaptively obtained based on the characteristics of the seismic signal in the time-frequency domain. The noise reduction threshold is determined based on the noise standard deviation. The noise standard deviation is denoted as The formula for the standard deviation of noise is: In the formula, M is the median of the wavelet transform coefficient modulus of the signal at the minimum scale. Let the denoising threshold be T, then the formula for the denoising threshold is: In the formula, ln(n) represents taking the natural logarithm with respect to the signal length n.

4. The seismic signal denoising method according to claim 2, characterized in that, The step of removing white noise from the time spectrum according to the denoising threshold to obtain the thresholded wavelet coefficient matrix includes: The wavelet coefficient matrix to be processed is determined based on the time spectrum. Based on the scale factor and time factor, the wavelet coefficient matrix to be processed is traversed, and the wavelet coefficients with a modulus value less than the denoising threshold are determined as the wavelet coefficients corresponding to white noise. By setting the wavelet coefficients corresponding to the white noise to zero, the wavelet coefficient matrix after thresholding is obtained.

5. A seismic signal denoising method according to claim 1, characterized in that, The step of performing median filtering on the impulse noise points in the first signal sequence to obtain a second signal sequence with impulse noise removed includes: Using the location of the impulse noise point as the center, extract the first segment of the signal with a window length of win; Sort the first segment of signal, and take the median and the values ​​that are adjacent to the median and are less than the median, and sum and average them to obtain the two-point mean; Replace the impulse noise points in the original acquired signal with the average of two points; The above steps are repeated for each impulse noise point in the first signal sequence, and a sliding window process is performed to obtain a second signal sequence with impulse noise removed.

6. A seismic signal denoising device employing the seismic signal denoising method according to any one of claims 1-5, characterized in that, include: The signal acquisition module is used to acquire the raw seismic signals. The white noise processing module is used to perform wavelet transform threshold denoising on the seismic signal to obtain a first signal sequence with white noise removed. The signal processing module is used to detect the first signal sequence, determine the position of each impulse noise in the first signal sequence, and mark it as an impulse noise point; The impulse noise processing module is used to perform median filtering on each impulse noise point in the first signal sequence to obtain a second signal sequence with impulse noise removed, which is then used as the denoised seismic signal sequence. The step of detecting the first signal sequence, determining the position of each impulse noise in the first signal sequence, and marking it as an impulse noise point includes: Determine the length of the detection window; Several detection points are set according to the first signal sequence; Based on the detection window length, each detection point is analyzed and processed sequentially, and the detection points that satisfy the preset discriminant are determined as impulse noise location points; Each impulse noise location point is marked as an impulse noise point; The step of sequentially analyzing and processing each detection point according to the detection window length, and determining the detection points that satisfy the preset discriminant as impulse noise location points, includes: Perform the following sliding process on each point to be detected sequentially until all impulse noise points are marked: Based on the detection window length and the position of the point to be detected, determine several neighboring signal sample points of the point to be detected within the detection window length range, and aggregate them into neighboring points; The differences between the point to be detected and each neighboring signal sample point are analyzed and processed to determine the number of differences that are greater than a preset difference threshold. Determine whether the ratio of the number of differences to the number of neighbor signal samples is greater than a preset percentage threshold; If so, mark the point to be detected as a pulse noise location point. The discriminant for determining the location of the impulse noise is as follows: In the formula, Points to be tested; For neighbor signal samples; These are the neighboring points of the point to be detected within the window length; A preset difference threshold is used to determine whether the difference between the point to be detected and its neighboring signal samples is large enough; The percentage threshold for determining the number of points with large differences between the point to be detected and its neighboring points; #[ ] represents the number of points that satisfy [ ].

7. An electronic device, characterized in that, The device includes a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that the processor, when executing the computer program, implements the steps of the denoising method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, When a computer program is executed by a processor, it implements the steps of the denoising method as described in any one of claims 1 to 5.

Images