A seismic signal random noise reduction processing method

Abstract

The application discloses a seismic signal random noise reduction processing method, comprising the following steps: S1, data preprocessing; S2, time-frequency analysis is performed to convert the time domain signal into a time-frequency domain representation; S3, in the time-frequency domain, a median filter noise recognition algorithm is applied by setting a threshold to identify and mark random noise; S4, the time-frequency representation is decomposed in multiple scales by using wavelet packet decomposition; S5, soft threshold or hard threshold processing is applied to the decomposed coefficients to reduce the influence of noise; S6, the processed coefficients are recombined, and the seismic signal is reconstructed through inverse transformation; S7, the reconstructed seismic signal is post-processed; S8, the noise reduction effect is evaluated by comparing the seismic signals before and after noise reduction, and using the signal-to-noise ratio and resolution index. The method is helpful for locating the source and distribution of noise, so that targeted processing can be performed; and the noise recognition algorithm can effectively separate noise from signals.

Description

Technical Field

[0001] This invention relates to a method for random noise reduction processing of seismic signals, belonging to the field of seismic signal processing technology. Background Technology

[0002] Seismic signal processing is an important branch of geophysics, primarily studying how to extract useful information from seismic data to understand the Earth's internal structure and tectonics. Seismic signal processing encompasses multiple aspects, such as data acquisition, preprocessing, noise suppression, signal separation, imaging, and inversion. The main objective of seismic signal processing is to extract the effective signals from seismic data to understand subsurface structures and tectonics. This requires the use of various signal processing methods, such as signal separation, imaging, and inversion. Signal separation involves separating the various signals mixed in seismic data for subsequent processing. Imaging and inversion convert seismic data into images of subsurface structures, thereby providing insights into these structures and tectonics. Seismic signal processing is a complex process that requires extensive knowledge and skills in various geophysical disciplines. With the advancement of science and technology, seismic signal processing methods and techniques are constantly improving, providing more possibilities for geophysical research.

[0003] However, common processing methods cannot identify the random noise characteristics in the signal during the noise reduction process, thus failing to locate the source and distribution of noise and cannot process it in a targeted manner. Summary of the Invention

[0004] The purpose of this invention is to provide a method for random noise reduction processing of seismic signals. This method identifies the characteristics of random noise in the signal through random noise identification, which helps to locate the source and distribution of noise, thus enabling targeted processing. Furthermore, the application of noise identification algorithms can effectively separate noise from the signal. This separation is the foundation of noise reduction processing, because only by accurately identifying noise can it be effectively suppressed, and its interference with the signal can be removed or reduced, thereby improving the clarity and readability of the signal. This provides valuable guidance for subsequent multi-scale decomposition and thresholding processing, enabling these processing steps to be performed more accurately and efficiently.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is: a method for random noise reduction processing of seismic signals, comprising the following steps:

[0006] S1. Data preprocessing: Preprocessing the raw seismic data;

[0007] S2. Perform time-frequency analysis: Apply time-frequency analysis to the preprocessed seismic data to convert the time-domain signal into a time-frequency domain representation;

[0008] S3. Random noise identification: In the time-frequency domain, random noise is identified and labeled by applying a medium-range filtering noise identification algorithm with a set threshold.

[0009] S4. Perform multi-scale decomposition: Use wavelet packet decomposition to perform multi-scale decomposition on the time-frequency representation;

[0010] S5. Perform thresholding: Apply soft or hard thresholding to the decomposed coefficients to reduce the impact of noise.

[0011] S6. Perform signal reconstruction: Recombine the processed coefficients and reconstruct the seismic signal through inverse transformation;

[0012] S7. Perform post-processing: Perform post-processing on the reconstructed seismic signal;

[0013] S8. Result verification: The noise reduction effect is evaluated by comparing the seismic signals before and after noise reduction, and by using signal-to-noise ratio and resolution metrics.

[0014] Furthermore, step S1 specifically includes the following steps:

[0015] S101. Data Import: Read the raw seismic data from the storage medium into the processing system;

[0016] S102. Bad sector detection and repair: Detect bad sectors in the data and repair them using data reconstruction technology;

[0017] S103. Outlier identification and processing: Identify outliers in the data and remove or replace them;

[0018] S104, Energy Balance: Perform energy balance processing on the data to compensate for differences in data energy;

[0019] S105. Normalization: Normalize the data to unify the data scale and facilitate subsequent processing and interpretation.

[0020] S106. Data quality control: Perform quality checks on the preprocessed data to ensure the effectiveness of the preprocessing steps.

[0021] Furthermore, the formula for energy balance is:

[0022]

[0023] Where x(t) is the original seismic data, A ref A(t) is the reference amplitude level, and A(t) is the amplitude at time t.

[0024] Furthermore, step S2 specifically includes the following steps:

[0025] S201. Select the short-time Fourier transform as the time-frequency analysis tool: By sliding a window function of fixed length across the signal and performing a Fourier transform on the signal within each window, the time-frequency representation of the signal can be obtained.

[0026] S202. Window function selection: Select an appropriate window function based on the signal characteristics and the required time-frequency resolution;

[0027] S203. Calculate STFT: Apply STFT to the preprocessed seismic data to calculate the frequency components within each time window;

[0028] S204, Time-Frequency Representation: The result of STFT is represented as a time-frequency spectrum.

[0029] Furthermore, step S3 specifically includes the following steps:

[0030] S301. Time-frequency analysis: Obtaining the time-frequency representation of seismic data using time-frequency analysis tools;

[0031] S302. Noise Characteristic Analysis: Analyze the characteristics of noise in the frequency spectrum.

[0032] S302, Setting a threshold: Set an appropriate threshold based on the noise characteristics to distinguish between signal and noise;

[0033] S304. Application of noise identification algorithm: Use median filtering identification technology to identify and label noise;

[0034] S305. Noise Suppression: Based on the identification results, suppress or remove the parts marked as noise.

[0035] Furthermore, step S4 specifically includes the following steps:

[0036] S401. Select wavelet packet decomposition as the multi-scale decomposition tool;

[0037] S402. Perform wavelet packet decomposition: Apply wavelet packet decomposition to the time-frequency representation to decompose the signal into wavelet packet coefficients at multiple scales;

[0038] S403. Select the number of decomposition levels: Determine the number of decomposition levels;

[0039] S404. Obtain wavelet packet coefficients: After decomposition, the wavelet packet coefficients at each scale represent the frequency components of the signal at that scale.

[0040] S405, Analysis Coefficients: Analyze wavelet packet coefficients to identify and separate useful signals from noise.

[0041] Furthermore, step S5 specifically includes the following steps:

[0042] S501. Select threshold processing method: Select soft threshold processing and hard threshold processing;

[0043] S502, Determine the threshold;

[0044] S503. Apply threshold processing: Apply the selected threshold method to the decomposed coefficients. For soft threshold processing, the coefficients are adjusted according to the soft threshold formula; for hard threshold processing, the coefficients are truncated according to the hard threshold formula.

[0045] S504, Coefficient Reconstruction: After thresholding, the processed coefficients are used to reconstruct the signal to obtain the denoised seismic signal.

[0046] Furthermore, step S6 specifically includes the following steps:

[0047] S601. Preparing the processed coefficients: After thresholding, a new set of coefficients is obtained, which represent the denoised version of the seismic signal.

[0048] S602, Perform inverse transformation: Apply inverse transformation to the processed coefficients;

[0049] S603, Reconstructed Signal: By inverse transforming, the coefficients are converted back to the time domain to obtain the denoised seismic signal;

[0050] S604. Result Verification: Perform a quality check on the reconstructed signal to ensure that the noise reduction process does not introduce new distortion or remove too many signal components.

[0051] Furthermore, step S7 specifically includes the following steps:

[0052] S701, Removing Boundary Effects: In the signal reconstruction process, boundary extension techniques are used to reduce the introduction of artifacts or boundary effects at the boundaries.

[0053] S702, Signal Enhancement: Apply enhancement techniques to the reconstructed seismic signal to improve its clarity and resolution;

[0054] S703. Data Quality Check: Perform a quality check on the post-processing results to ensure the effectiveness of the processing steps and to check whether new noise or distortion has been introduced.

[0055] S704. Output Results: Output the processed seismic signal for further seismic data interpretation.

[0056] Furthermore, in step S8, the formula for calculating the signal-to-noise ratio is:

[0057]

[0058] Where x(t) is the original signal and x^(t) is the denoised signal.

[0059] The beneficial effects of this invention are:

[0060] (1) In this invention, random noise identification can identify the characteristics of random noise in the signal, which helps to locate the source and distribution of noise, thereby enabling targeted processing; and the application of noise identification algorithms can effectively separate noise from the signal. This separation is the basis of noise reduction processing, because only by accurately identifying noise can it be effectively suppressed, and the interference of this noise on the signal can be removed or reduced, thereby improving the clarity and readability of the signal, providing valuable guidance for subsequent multi-scale decomposition and threshold processing, and enabling these processing steps to be performed more accurately and efficiently.

[0061] (2) In this invention, thresholding significantly reduces the noise component in the decomposed wavelet packet coefficients. This process removes noise while preserving useful signal information, suppressing noise and retaining key signal characteristics such as amplitude and frequency. This helps preserve the original signal properties, making the reconstructed signal more realistic and reliable, thus significantly improving the signal-to-noise ratio and signal quality. This aids in subsequent seismic data interpretation and oil and gas exploration; and compared to other noise reduction methods, it has lower computational complexity. This makes the processing more efficient, saves computational resources, and allows for adjustments based on the specific signal and noise characteristics, exhibiting good adaptability, making the processing method more flexible and customizable. In summary, random noise identification and thresholding have significant beneficial effects on random noise reduction methods for seismic signals. These two steps effectively improve signal quality and signal-to-noise ratio, providing a reliable data foundation for subsequent processing and interpretation. Attached Figure Description

[0062] Figure 1 This is a flowchart of a random noise reduction method for seismic signals. Detailed Implementation

[0063] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0064] like Figure 1 A method for random noise reduction processing of seismic signals, the method comprising the following steps:

[0065] S1: Perform data preprocessing. Preprocess the raw seismic data, including removing outliers, bad paths, and other anomalies, and performing energy balancing and normalization to ensure data quality and lay the foundation for subsequent processing.

[0066] S2: Perform time-frequency analysis, apply time-frequency analysis techniques such as short-time Fourier transform (STFT) to the preprocessed seismic data to convert the time-domain signal into a time-frequency domain representation;

[0067] S3: Perform random noise identification. In the time-frequency domain, random noise is identified and labeled by setting a threshold or applying noise identification algorithms (such as median filtering, K-means clustering, etc.). Random noise typically refers to noise that cannot be predicted by physical models and usually manifests in the high-frequency components of a signal.

[0068] S4: Perform multi-scale decomposition on the time-frequency representation, for example, using wavelet packet decomposition;

[0069] S5: Perform thresholding. Apply soft or hard thresholding to the decomposed coefficients to reduce the impact of noise.

[0070] S6: Perform signal reconstruction, recombine the processed coefficients, and reconstruct the seismic signal through inverse transformation;

[0071] S7: Post-processing is performed on the reconstructed seismic signal, including removing boundary effects that may be introduced during the reconstruction process and enhancing the signal.

[0072] S8: Perform result verification by comparing seismic signals before and after noise reduction, and evaluating the noise reduction effect using metrics such as signal-to-noise ratio (SNR) and resolution.

[0073] Step S1 specifically includes the following steps:

[0074] S101. Data Import: Read the raw seismic data from the storage medium into the processing system.

[0075] S102. Bad Sector Detection and Repair: Detect bad sectors in the data (such as data loss or anomalies caused by cable faults, sensor problems, etc.). Repair bad sectors using interpolation or other data reconstruction techniques.

[0076] S103. Outlier Identification and Processing: Identify outliers (isolated points, outliers) in the data, which may be caused by environmental interference, instrument malfunction, etc. Remove or replace outliers using median filtering, least squares method, or other smoothing techniques.

[0077] S104. Energy Balancing: Energy balancing is performed on the data to compensate for differences in data energy caused by factors such as geological conditions and acquisition environment. This can be achieved through Automatic Gain Control (AGC). The purpose of AGC is to keep the signal amplitude approximately constant, making seismic data from different regions comparable.

[0078] S105. Normalization: Normalize the data to unify the data scale, facilitating subsequent processing and interpretation. Methods such as maximum value normalization and standard deviation normalization can be used.

[0079] S106. Data Quality Control: Perform quality checks on the preprocessed data to ensure the effectiveness of the preprocessing steps. This can be done through methods such as visualization, statistical analysis, and manual review.

[0080] The formula for energy balance (AGC) is:

[0081]

[0082] Where x(t) is the original seismic data, A ref A(t) is the reference amplitude level, and A(t) is the amplitude at time t.

[0083] Step S2 specifically includes the following steps:

[0084] S201. Select a time-frequency analysis tool: Short-Time Fourier Transform (STFT) is one of the most commonly used time-frequency analysis tools. It obtains the time-frequency representation of a signal by sliding a fixed-length window function across the signal and performing a Fourier transform on the signal within each window.

[0085] S202. Window Function Selection: Choose an appropriate window function, such as a Hamming window, Hanning window, or Gaussian window. The choice of window function will affect the time-frequency resolution and usually needs to be determined based on the characteristics of the signal and the required time-frequency resolution.

[0086] S203. Calculate STFT: Apply STFT to the preprocessed seismic data to calculate the frequency components within each time window.

[0087] S204. Time-Frequency Representation: The results of STFT are represented as a time spectrum, usually displayed as a color intensity graph, where time is along the horizontal axis, frequency is along the vertical axis, and color intensity represents amplitude or energy.

[0088] Step S3 specifically includes the following steps:

[0089] S301. Time-frequency analysis: Obtain the time-frequency representation of seismic data using STFT or other time-frequency analysis tools.

[0090] S302. Noise Characteristic Analysis: Analyze the characteristics of noise in the frequency spectrum, such as frequency range and amplitude distribution.

[0091] S303. Set Threshold: Set an appropriate threshold based on noise characteristics to distinguish between signal and noise. The threshold can be determined through experience, statistical analysis, or automatic optimization algorithms.

[0092] S304. Application of Noise Identification Algorithms: Use median filtering, K-means clustering, pattern recognition, or other noise identification techniques to identify and label noise. Median filtering is a commonly used nonlinear filtering technique that removes noise by sliding a window across the time spectrum and replacing the center value with the median value of the data within the window. K-means clustering is an unsupervised learning method that can divide data points into K categories, where noise can be a separate category.

[0093] S305. Noise Suppression: Based on the identification results, suppress or remove the parts marked as noise.

[0094] Step S4 specifically includes the following steps:

[0095] S401. Select a multi-scale decomposition tool: Wavelet packet decomposition is a flexible multi-scale decomposition tool that allows signals to be decomposed at different scales and can provide finer-grained frequency information.

[0096] S402. Perform wavelet packet decomposition: Apply wavelet packet decomposition to the time-frequency representation to decompose the signal into wavelet packet coefficients at multiple scales.

[0097] S403. Choosing the Decomposition Level: Determine the level of decomposition, which will affect the level of detail in the decomposition. The higher the level, the more detailed the decomposition, but the higher the computational complexity.

[0098] S404. Obtain wavelet packet coefficients: After decomposition, the wavelet packet coefficients at each scale represent the frequency components of the signal at that scale.

[0099] S405. Analysis of Coefficients: Analyze the wavelet packet coefficients to identify and separate useful signals from noise. Typically, the high-frequency portion of a signal may contain noise, while the low-frequency portion contains the valid signal.

[0100] Step S5 specifically includes the following steps:

[0101] S501. Selecting a threshold processing method: There are generally two methods for threshold processing: soft thresholding and hard thresholding. Soft thresholding sets coefficients smaller than the threshold to zero, while hard thresholding truncates coefficients smaller than the threshold.

[0102] S502. Determining the Threshold: The choice of threshold is crucial to the noise reduction effect. The threshold can be determined by methods such as general thresholding, SURE thresholding, and minimizing maximum likelihood estimation.

[0103] S503. Apply Thresholding: Apply the selected thresholding method to the decomposed coefficients. For soft thresholding, the coefficients are adjusted according to the soft thresholding formula; for hard thresholding, the coefficients are truncated according to the hard thresholding formula.

[0104] S504. Coefficient Reconstruction: After thresholding, the processed coefficients are used to reconstruct the signal to obtain the denoised seismic signal.

[0105] Step S6 specifically includes the following steps:

[0106] S601. Preparing the processed coefficients: After thresholding, a new set of coefficients is obtained, which represent the denoised version of the seismic signal.

[0107] S602. Perform inverse transform: Apply the inverse transform to the processed coefficients. For example, if wavelet packet decomposition is used, the inverse wavelet packet transform (composition transform) needs to be applied.

[0108] S603. Reconstructing the signal: By inverse transformation, the coefficients are converted back to the time domain to obtain the denoised seismic signal.

[0109] S604. Result Verification: Perform a quality check on the reconstructed signal to ensure that the noise reduction process does not introduce new distortion or remove too many signal components.

[0110] Step S7 specifically includes the following steps:

[0111] S701. Removing Boundary Effects: During signal reconstruction, especially when using wavelet packets or other multi-scale decomposition techniques, artifacts or boundary effects may be introduced at the signal boundaries. These effects can be reduced using boundary continuation techniques (such as symmetric continuation, periodic continuation, etc.).

[0112] S702. Signal Enhancement: Enhancement techniques are applied to the reconstructed seismic signal to improve its clarity and resolution. Commonly used signal enhancement techniques include amplitude enhancement, frequency domain filtering, and adaptive filtering.

[0113] S703. Data Quality Check: Perform a quality check on the post-processing results to ensure the effectiveness of the processing steps and to check whether new noise or distortion has been introduced.

[0114] S704. Output Results: Outputs the processed seismic signal for further seismic data interpretation or other applications.

[0115] In step S8, the formula for calculating the signal-to-noise ratio is:

[0116]

[0117] Where x(t) is the original signal and x^(t) is the denoised signal.

[0118] The working principle of this invention is as follows: 1. By performing random noise identification, the characteristics of random noise in the signal can be identified, which helps to locate the source and distribution of noise, thus enabling targeted processing. Furthermore, the noise identification algorithm can effectively separate noise from the signal. This separation is the foundation of noise reduction processing, because only by accurately identifying noise can it be effectively suppressed, and its interference with the signal can be removed or reduced, thereby improving signal clarity and readability. This provides valuable guidance for subsequent multi-scale decomposition and thresholding processing, enabling these processing steps to be performed more accurately and efficiently. 2. Through thresholding processing, the noise component in the wavelet packet coefficients after decomposition can be significantly reduced. This processing can remove noise while retaining useful information in the signal. It not only suppresses noise but also preserves the main characteristics of the signal, such as amplitude and frequency. This helps to preserve the original characteristics of the signal, making the reconstructed signal more realistic and reliable; thus significantly improving the signal-to-noise ratio and enhancing signal quality. This is helpful for subsequent seismic data interpretation and oil and gas exploration; and compared to other noise reduction methods, it has lower computational complexity. This makes the processing more efficient, saves computational resources, and allows for adjustments based on the specific signal conditions and noise characteristics, exhibiting good adaptability and making the processing method more flexible and customizable. In summary, random noise identification and thresholding have significant beneficial effects on random noise reduction methods for seismic signals. These two steps can effectively improve signal quality and signal-to-noise ratio, providing a reliable data foundation for subsequent processing and interpretation.

[0119] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the scope of protection of the present invention in any way, and all technical solutions obtained by equivalent substitution or other means fall within the scope of protection of the present invention. Parts not covered in this invention are the same as or can be implemented using existing technology.

Claims

1. A method for random noise reduction processing of seismic signals, characterized in that, Includes the following steps: S1. Data preprocessing: Preprocessing the raw seismic data; S2. Perform time-frequency analysis: Apply time-frequency analysis to the preprocessed seismic data to convert the time-domain signal into a time-frequency domain representation; S3. Random noise identification: In the time-frequency domain, random noise is identified and labeled by applying a medium-range filtering noise identification algorithm with a set threshold. S4. Perform multi-scale decomposition: Use wavelet packet decomposition to perform multi-scale decomposition on the time-frequency representation; S5. Perform thresholding: Apply soft or hard thresholding to the decomposed coefficients to reduce the impact of noise. S6. Perform signal reconstruction: Recombine the processed coefficients and reconstruct the seismic signal through inverse transformation; S7. Perform post-processing: Perform post-processing on the reconstructed seismic signal; S8. Result verification: The noise reduction effect is evaluated by comparing the seismic signals before and after noise reduction, and by using signal-to-noise ratio and resolution metrics.

2. The method for random noise reduction processing of seismic signals according to claim 1, characterized in that, Step S1 specifically includes the following steps: S101. Data Import: Read the raw seismic data from the storage medium into the processing system; S102. Bad sector detection and repair: Detect bad sectors in the data and repair them using data reconstruction technology; S103. Outlier identification and processing: Identify outliers in the data and remove or replace them; S104, Energy Balance: Perform energy balance processing on the data to compensate for differences in data energy; S105. Normalization: Normalize the data to unify the data scale and facilitate subsequent processing and interpretation. S106. Data quality control: Perform quality checks on the preprocessed data to ensure the effectiveness of the preprocessing steps.

3. The method for random noise reduction processing of seismic signals according to claim 2, characterized in that, The formula for energy balance is: Where x(t) is the original seismic data, A ref A(t) is the reference amplitude level, and A(t) is the amplitude at time t.

4. The method for random noise reduction processing of seismic signals according to claim 1, characterized in that, Step S2 specifically includes the following steps: S201. Select the short-time Fourier transform as the time-frequency analysis tool: By sliding a window function of fixed length across the signal and performing a Fourier transform on the signal within each window, the time-frequency representation of the signal can be obtained. S202. Window function selection: Select an appropriate window function based on the signal characteristics and the required time-frequency resolution; S203. Calculate STFT: Apply STFT to the preprocessed seismic data to calculate the frequency components within each time window; S204, Time-Frequency Representation: The result of STFT is represented as a time-frequency spectrum.

5. The method for random noise reduction processing of seismic signals according to claim 1, characterized in that, Step S3 specifically includes the following steps: S301. Time-frequency analysis: Obtaining the time-frequency representation of seismic data using time-frequency analysis tools; S302. Noise Characteristic Analysis: Analyze the characteristics of noise in the frequency spectrum. S302, Setting a threshold: Set an appropriate threshold based on the noise characteristics to distinguish between signal and noise; S304. Application of noise identification algorithm: Use median filtering identification technology to identify and label noise; S305. Noise Suppression: Based on the identification results, suppress or remove the parts marked as noise.

6. The method for random noise reduction processing of seismic signals according to claim 1, characterized in that, Step S4 specifically includes the following steps: S401. Select wavelet packet decomposition as the multi-scale decomposition tool; S402. Perform wavelet packet decomposition: Apply wavelet packet decomposition to the time-frequency representation to decompose the signal into wavelet packet coefficients at multiple scales; S403. Select the number of decomposition levels: Determine the number of decomposition levels; S404. Obtain wavelet packet coefficients: After decomposition, the wavelet packet coefficients at each scale represent the frequency components of the signal at that scale. S405, Analysis Coefficients: Analyze wavelet packet coefficients to identify and separate useful signals from noise.

7. The method for random noise reduction processing of seismic signals according to claim 1, characterized in that, Step S5 specifically includes the following steps: S501. Select threshold processing method: Select soft threshold processing and hard threshold processing; S502, Determine the threshold; S503. Apply threshold processing: Apply the selected threshold method to the decomposed coefficients. For soft threshold processing, the coefficients are adjusted according to the soft threshold formula; for hard threshold processing, the coefficients are truncated according to the hard threshold formula. S504, Coefficient Reconstruction: After thresholding, the processed coefficients are used to reconstruct the signal to obtain the denoised seismic signal.

8. The method for random noise reduction processing of seismic signals according to claim 1, characterized in that, Step S6 specifically includes the following steps: S601. Preparing the processed coefficients: After thresholding, a new set of coefficients is obtained, which represent the denoised version of the seismic signal. S602, Perform inverse transformation: Apply inverse transformation to the processed coefficients; S603, Reconstructed Signal: By inverse transforming, the coefficients are converted back to the time domain to obtain the denoised seismic signal; S604. Result Verification: Perform a quality check on the reconstructed signal to ensure that the noise reduction process does not introduce new distortion or remove too many signal components.

9. The method for random noise reduction processing of seismic signals according to claim 1, characterized in that, Step S7 specifically includes the following steps: S701, Removing Boundary Effects: In the signal reconstruction process, boundary extension techniques are used to reduce the introduction of artifacts or boundary effects at the boundaries. S702, Signal Enhancement: Apply enhancement techniques to the reconstructed seismic signal to improve its clarity and resolution; S703. Data Quality Check: Perform a quality check on the post-processing results to ensure the effectiveness of the processing steps and to check whether new noise or distortion has been introduced. S704. Output Results: Output the processed seismic signal for further seismic data interpretation.

10. The method for random noise reduction processing of seismic signals according to claim 1, characterized in that, In step S8, the formula for calculating the signal-to-noise ratio is: Where x(t) is the original signal and x^(t) is the denoised signal.

Images