A seismic signal denoising method based on local coherence constraint improved Radon transform

Abstract

This invention provides a seismic signal denoising method based on an improved Radon transform with local coherence constraints, belonging to the field of seismic signal denoising technology. The method first reads the original noisy seismic signal data and performs a Radon transform to obtain the -p domain representation of the noisy seismic signal; then, it performs time correction and alignment processing on the original seismic signal along the corresponding slowness; a sliding window is set along the spatial direction on the aligned seismic signal data, and local similarity coefficients are calculated at each location, with the maximum local similarity coefficient extracted; the maximum local similarity coefficient is used as a local coherence constraint factor to perform soft weighting processing on the Radon transform values ​​in the -p domain, obtaining weighted Radon transform values ​​in the -p domain; the weighted Radon transform values ​​in the -p domain are then inversely transformed to obtain the denoised and reconstructed seismic signal. This method achieves effective suppression of low-coherence noise and faithful preservation of high-coherence effective signals in seismic signals.

Description

Technical Field

[0001] This invention relates to the field of seismic signal denoising technology, and in particular to a seismic signal denoising method based on an improved Radon transform with local coherence constraints. Background Technology

[0002] Improving the signal-to-noise ratio is one of the main tasks in seismic data processing and interpretation. Due to factors such as complex geological conditions, construction environment, and observation systems, actual acquired seismic signals often contain a large amount of random noise and coherent interference signals, severely affecting the identification of valid seismic events and subsequent imaging accuracy. Therefore, researching efficient and stable seismic signal denoising methods is of great significance for improving seismic data quality and imaging resolution.

[0003] Currently, in practical seismic data processing, commonly used signal processing and denoising techniques mainly include time-domain methods, frequency-domain methods, time-frequency analysis methods, and transform-domain methods. Among them, the Radon transform (… The -p) and its Radon transform form in seismic data processing have been widely used in seismic event separation and feature extraction because they can map wavefields with different propagation characteristics to the parameter domain.

[0004] Chinese invention patent application 202411359919.X discloses a seismic surface wave array tomography method. This method is based on continuously recorded data from seismic stations. It calculates the cross-correlation function between station pairs and processes the cross-correlation data within different subarrays using a linear Radon transform to extract the dispersion curves of each subarray. This is then combined with array tomography to obtain the phase velocity structure of the study area. While this invention utilizes the mapping properties of the Radon transform in the parameter domain to improve the spatial resolution of array processing and imaging, it primarily focuses on surface wave dispersion feature extraction and imaging applications, without delving into the local distinction between noise and effective seismic events within the Radon transform domain.

[0005] Chinese invention patent application 201911268666.4 discloses a sample entropy thresholding method for denoising microseismic signals based on integrated empirical mode decomposition (EMD). This method obtains the intrinsic mode functions (EMFs) of different frequency components by performing EMD on the microseismic signal, and then reconstructs the signal based on the sample entropy index, selecting mode components that match the characteristics of the microseismic signal, thereby achieving noise suppression and signal-to-noise ratio (SNR) improvement. While this type of method improves the denoising effect of microseismic signals to some extent, its processing mainly relies on EMD and time-domain statistical feature analysis, without utilizing the event slope characteristics in the Radon transform domain, and has limited ability to distinguish different events under complex wavefield conditions.

[0006] Chinese invention patent application 202410107608.8 discloses an array-based geomagnetic field signal acquisition system and method. This invention addresses the issue at the signal acquisition source, suppressing interference signals from non-target directions through array deployment and electromagnetic shielding. It also combines signal coherence analysis to remove interference from the acquired electromagnetic signals, thereby improving signal purity. The invention proposes starting from the "signal source," employing an array-distributed magnetic field signal acquisition device. First, electromagnetic shielding is used to block signals from the left, right, and below the magnetic sensor, receiving only signals from above the sensor, thus initially achieving shielding of "non-planar electromagnetic signals" and improving signal purity. Then, signal coherence analysis is used to remove electromagnetic interference signals, achieving signal-to-noise separation. This scheme embodies the idea of ​​using coherence characteristics to distinguish effective signals from interference signals; however, it is mainly applied at the electromagnetic signal acquisition system level and does not incorporate coherence constraints into the Radon transform domain processing of seismic signals, nor does it provide refined constraints for different events within the seismic signals.

[0007] The above analysis shows that existing technologies have made some progress in applications of Radon transform, mode decomposition denoising, and coherence analysis, but still have the following shortcomings: On the one hand, existing Radon transform-based methods mostly employ global processing or energy criteria, making it difficult to effectively distinguish between local short-coherence noise and effective seismic events in the Radon transform domain; on the other hand, existing coherence or threshold-based methods are mostly applied to the time domain or other transform domains, and have not yet formed a system specifically for... A joint constraint mechanism for event-level features in the -p domain. Therefore, it is necessary to propose a mechanism that can introduce local coherence constraints in the Radon transform domain, based on local similarity coefficients. -p domain results are adaptively soft-weighted to improve denoising accuracy and stability under complex wavefield conditions. Summary of the Invention

[0008] The technical problem to be solved by this invention is to address the shortcomings of the existing technology by providing a seismic signal denoising method based on locally coherent constraint-based improved Radon transform, which addresses the deficiencies of existing Radon transform (…). The denoising process of the -p domain transform often uses global thresholds or energy criteria, which makes it difficult to effectively distinguish between local short coherent noise and effective seismic events, and easily weakens the effective signal under complex wavefield conditions. By introducing a local coherence constraint mechanism in the Radon transform domain and performing adaptive weighting processing on the transform results, we can achieve effective suppression of low coherent noise in seismic signals and fidelity preservation of high coherent effective signals, thereby improving the denoising effect and processing stability.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: On the one hand, this invention provides a seismic signal denoising method based on locally coherent constraint-improved Radon transform, comprising the following steps: Read raw, noisy seismic signal data; Performing a Radon transform on the original noisy seismic signal yields the noisy seismic signal. -p field representation; based on The propagation slowness parameter p of different seismic events in the -p domain is used to perform time correction and alignment processing on the original seismic signal along the corresponding slowness. A sliding window is set along the spatial direction on the aligned seismic signal data, and the local similarity coefficient is calculated at each location, and the maximum local similarity coefficient is extracted. Using the maximum local similarity coefficient as a local coherence constraint factor The Radon transform values ​​within the -p domain are then soft-weighted to obtain the weighted result. - Values ​​of each Radon transform within the p domain; For weighted The -p domain Grardon transform is used to perform an inverse transform to obtain the denoised and reconstructed seismic signal.

[0010] Furthermore, the original noisy seismic signal data is represented as follows: ,in, This indicates the trace number of each receiver channel in the seismic data or the spatial sampling location of the seismic record. Indicates the time sampling point.

[0011] Furthermore, based on The propagation slowness parameter p for different seismic events in the -p domain is used to perform time correction and alignment processing on the original seismic signal along the corresponding slowness, so that seismic events that meet the slowness condition are aligned in the spatial direction. Its expression is: ; in, To adjust the original seismic signal according to the time correction relationship t= p, the original seismic signal is adjusted for the slowness parameter p of the seismic event propagation. The signal obtained after alignment is used to characterize the spatial consistency of the data in each channel under this propagation slowness condition. This is the intercept time.

[0012] Furthermore, the formula for calculating the local similarity coefficient is as follows: ; in, For the first Within a sliding window -p domain Local similarity coefficient at location, Indicates the first A sliding window in space, This indicates the number of valid sample points participating in the calculation within the current window. As a stabilizing factor; During the movement of the sliding window, the same The maximum local similarity coefficient is obtained by taking the maximum value of the local similarity coefficients calculated from different windows at a given location. The calculation formula is as follows: .

[0013] Furthermore, the weighted The formulas for calculating the values ​​of each Radon transform within the -p domain are shown below: ; in, The original seismic signal is transformed by Radon transform and mapped to a coordinate system with slowness and intercept time. -p domain The value of the Radon transform at the location, For weighted The values ​​of each Radon transform within the -p domain The weighting factor is calculated using the following formula: ; In the formula, This represents the lower limit of the local similarity coefficient. This is the coherence weighting index.

[0014] Furthermore, the soft-weighted processing The inverse transform of the -p domain Radon transform yields the denoised and reconstructed seismic signal, as shown in the following formula: .

[0015] Furthermore, when the maximum local similarity coefficient Less than the preset threshold At that time, for the corresponding Additional attenuation constraints are applied at the -p domain location to further reduce the energy of low-coherence background noise.

[0016] Furthermore, the method employs a multiplicative attenuation method for the corresponding... - Additional attenuation constraints are applied to the p-domain location.

[0017] Secondly, this application proposes a computer-readable storage medium storing executable instructions that, when executed, cause a processor to perform the seismic signal denoising method based on the improved Radon transform with local coherence constraints.

[0018] Thirdly, this application proposes a computer program product, including a computer program or instructions that, when executed by a processor, implement the seismic signal denoising method based on the improved Radon transform with local coherence constraints.

[0019] The beneficial effects of the above technical solution are as follows: The seismic signal denoising method based on local coherence constraint and improved Radon transform provided by the present invention (1) can effectively distinguish between high coherence effective seismic events and low coherence noise by introducing local coherence constraint in the Radon transform domain; (2) adopting a soft weighting method based on local similarity coefficient avoids excessive weakening of effective signals by the traditional global hard threshold method; (3) by setting a sliding window along the spatial direction and extracting the maximum local similarity coefficient, the influence of a single fixed window position on the local coherence evaluation results is reduced, and the adaptability to local bending events, non-strict global continuous events and continuous events within a certain trace range is improved; (4) the method of the present invention has a clear process and is applicable to various processing scenarios such as synthetic seismic data, complex model data and measured seismic data, and has good versatility and engineering application value. Attached Figure Description

[0020] Figure 1 A flowchart of a seismic signal denoising method based on locally coherent constraint-improved Radon transform provided in an embodiment of the present invention; Figure 2 This is a noisy seismic record image of the synthetic seismic data provided in the embodiments of the present invention; Figure 3 The original seismic record image with noise from Marmousi provided for embodiments of the present invention; Figure 4 Synthetic seismic data provided for embodiments of the present invention -p domain conventional Radon transform result diagram; Figure 5 Introducing local coherence constraints into the synthetic data provided in the embodiments of the present invention -p domain result graph; Figure 6 The noise-adding marmousi provided in the embodiments of the present invention -p domain conventional Radon transform result diagram; Figure 7 The local coherence constraint is introduced for the noisy marmousi provided in the embodiments of the present invention. -p domain result graph; Figure 8 An inverse transformation diagram of the results of conventional methods on the synthetic seismic data provided in the embodiments of the present invention; Figure 9 An inverse transformation diagram of the results of the improved method based on the synthetic seismic data provided in the embodiments of the present invention; Figure 10 The inverse transformation diagram of the conventional result with noisy Marmousi provided in the embodiments of the present invention; Figure 11 The inverse transformation diagram of the improved result of the noisy Marmousi algorithm provided in the embodiments of the present invention. Detailed Implementation

[0021] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0022] Example 1: In this embodiment, a seismic signal denoising method based on locally coherent constraint-improved Radon transform is described, such as... Figure 1 As shown, it includes the following steps: Step 1: Read the raw, noisy seismic signal data; In this embodiment, the raw, noisy seismic signal data is first read. This seismic signal can be synthetic seismic data, seismic data from a noisy complex model, or measured seismic data. Its time-range domain representation is denoted as... ,in, This indicates the trace number of each receiver channel in the seismic data or the spatial sampling location of the seismic record. Indicates the time sampling point. For example... Figure 2 and Figure 3 The diagram shown is a schematic of the original noisy seismic record. It can be seen that effective seismic events, random noise, and coherent noise are intertwined and distributed in the time-range domain, making it difficult to effectively distinguish them directly in the time domain.

[0023] Step 2: Perform a Radon transform on the original noisy seismic signal to obtain the noisy seismic signal. -p field representation; The original noisy seismic signal was subjected to Radon transform. -p transform), mapping the time-range domain signal to... The -p field, its mathematical expression is: ; in, This is a parameter representing the slowness of earthquake event propagation. Intercept time; The original seismic signal is transformed by Radon transform and mapped to a coordinate system with slowness and intercept time. -p domain The Lardon transform value at the location. All Lardon transform values. Together they constitute the original seismic signal. Transformation results in the -p domain; For synthetic seismic data, its conventional Radon transform The result of the -p field is as follows Figure 4 As shown; for noisy and complex model seismic data, the conventional Radon transform result is as follows. Figure 6 As shown. By Figure 4 and Figure 6 It can be seen that, without the introduction of local constraints, Within the -p domain, the effective seismic events and noise energy distribution are relatively dispersed, with some short-coherent noise... The -p domain still exhibits characteristics similar to linear events.

[0024] Step 3: Perform time correction and alignment processing on the original seismic signal according to different slowness parameters; against The propagation slowness parameter p for different seismic events in the -p domain is used to perform time correction and alignment processing on the original seismic signal along the corresponding slowness, so that seismic events that meet the slowness condition are aligned in the spatial direction. Its expression is: ; in, To adjust the propagation slowness parameter p, the original seismic signal is corrected according to the time correction relationship t= The signal obtained after alignment is used to characterize the consistency of each channel of data in the spatial direction under the condition of slow propagation. Through the above processing, under the correct slowness condition, effective seismic events show high consistency in spatial direction, while random noise and interference events that do not match the current slowness still show low consistency distribution, providing a basis for subsequent local similarity analysis.

[0025] Step 4: Set a sliding window along the spatial direction on the aligned seismic signal data, calculate the local similarity coefficient, and extract the maximum local similarity coefficient; In the aligned seismic signal, a sliding window is set along the spatial direction, and the local similarity coefficient is calculated at each location, denoted as . The calculation formula is as follows: ; in, Indicates the first A sliding window in space, This indicates the number of valid sample points participating in the calculation within the current window. It is a stabilizing factor.

[0026] During the movement of the sliding window, the same The maximum local similarity coefficient is obtained by taking the maximum value of the local similarity coefficients calculated from different windows at a given location. ; The maximum local similarity coefficient is used to characterize The maximum local coherence property at the corresponding position in the -p domain.

[0027] Step 5: Use the maximum local similarity coefficient as a local coherence constraint factor to... The values ​​of each Radon transform in the -p domain are soft-weighted to enhance and preserve high-coherence effective seismic events and suppress low-coherence noise energy. The calculated maximum local similarity coefficient is used as a constraint. The -p domain transformation result is then subjected to soft weighting of the Radon transform coefficients to obtain the weighted result. -p domain Gerardon transform values : ; in, The weighting factor is expressed as: ; in, This represents the lower limit of the local similarity coefficient. This is the coherence weighting index. In this weighting process, when... When the value is large, the corresponding weighting factor is large, and highly coherent effective seismic events are in The energy in the -p domain is enhanced; when When the value is smaller, the weighting factor at the corresponding location decreases, and low-coherence noise and spurious energy are suppressed. By setting a lower limit for the local similarity coefficient, it is possible to avoid excessive suppression of low-coherence regions, which could lead to the complete loss of weak and effective signals.

[0028] For synthetic seismic data, after introducing local coherence constraints The result of the -p field is as follows Figure 5 As shown; for noisy and complex model seismic data, the introduction of local coherence constraints... The result of the -p field is as follows Figure 7 As shown. With Figure 4 and Figure 6 In comparison, it can be seen that In the -p domain, the energy of highly coherent events is more concentrated, while the energy of low-coherent noise is effectively suppressed.

[0029] Furthermore, in an alternative implementation, when the maximum local similarity coefficient Less than the preset threshold At the same time, it is also possible to target the corresponding Additional attenuation constraints are applied to the -p domain location to further reduce the energy of low-coherence background noise; when it is not less than the gate threshold, the soft-weighted result remains unchanged.

[0030] In this embodiment, corresponding Applying additional attenuation constraints to the -p domain can be achieved using multiplicative attenuation, for example, on a soft-weighted domain. The result in the -p domain is then multiplied by an attenuation coefficient less than 1; when the maximum local similarity coefficient is lower than the preset gating threshold, additional attenuation is applied, and when the maximum local similarity coefficient is not lower than the preset gating threshold, the soft-weighted result remains unchanged.

[0031] Step 6: Processing the soft-weighted data - The Lardon transform value in the p domain is used for inverse transform to achieve denoising and reconstruction of the seismic signal; After local coherence constraint processing The -p domain data is subjected to inverse Lardon transform, and the reconstructed denoised seismic signal is obtained, the expression of which is: ; For discrete seismic signal data, its inverse transform form can be expressed as: ; in, Indicates spatial location With time sampling points The denoised seismic signal obtained from the reconstruction. Indicates the first A spatial sampling location or receiving channel location. Indicates the first Each time sampling point; Indicates the first One slow sampling point, Indicates the total number of slow-speed samples. Indicates the slow sampling interval; Indicates slow speed Under the condition of soft weighting with local coherence constraints, -p domain Radon transform values, where For the corresponding intercept time variable. This formula represents the discrete summation of the weighted Radon transform values ​​at all slow sampling points, thereby achieving the result from... Reconstruction using the inverse Lardon transform from the p-domain to the time-range domain. For synthetic seismic data, the difference in denoising effect before and after is as follows: Figure 8 and Figure 9 As shown; for noisy and complex seismic model data, the comparison of the effects before and after denoising is as follows: Figure 10 and Figure 11 As shown in the figure, after reconstruction using the inverse Lardon transform, the denoising results significantly reduce noise interference while maintaining the effective amplitude and morphological characteristics of seismic events.

[0032] In practical application, this invention provides an improved Radon transform seismic signal denoising method based on local coherence constraints, which processes random noise and short-coherence linear noise. Analysis and calculation results show that in conventional Radon transform processing, The effective seismic events and noise energy distribution within the -p domain are relatively dispersed, with some noise... The -p domain still exhibits characteristics similar to linear events, such as... Figure 4 and Figure 6 As shown, using a conventional Radon transform followed by a direct inverse transform can easily result in residual noise or attenuation of the effective signal. Furthermore, when processing using only a global threshold or a simple energy criterion... In the -p domain, low-coherence noise and high-coherence valid events are difficult to distinguish effectively, easily leading to the loss of local details in valid seismic events. In contrast, such as Figure 5 and Figure 7 As shown, the present invention, through Local coherence constraints are introduced within the -p domain, and the low-coherence noise is soft-weighted and suppressed using the maximum local similarity coefficient, so that the energy of high-coherence seismic events is reduced. The -p domain is more concentrated. (Combined) Figure 8 , Figure 9 as well as Figure 10 , Figure 11 It can be seen that after reconstruction by inverse Radon transform, the denoising results effectively reduce background noise and stray interference while maintaining the continuity of effective seismic events and the main amplitude characteristics.

[0033] The improved Radon transform seismic signal denoising method based on local coherence constraints described in this invention combines local similarity analysis with... The -p domain soft weighting constraint enables a fine distinction between noise and valid events in seismic signals, making the denoising process smoother and more targeted. Under different types of seismic data conditions, this method demonstrates good stability and applicability, especially in cases with complex noise components and weak coherence of valid events, maintaining good denoising performance and exhibiting significant technical advantages.

[0034] Step 7: Output the denoised seismic signal results.

[0035] Example 2: This embodiment proposes a computer-readable storage medium that stores executable instructions. When these instructions are executed, if they are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium.

[0036] The computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the improved Radon transform seismic signal denoising method based on local coherence constraints described in the various embodiments of this application.

[0037] The aforementioned storage media include: flash memory, hard disks, multimedia cards, card-type memory (e.g., SD (Secure Digital Memory Card) or DX (Memory Data Register, MDR) memory), random access memory (RAM), static random-access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic storage, disks, optical discs, servers, APP (Application) app stores, and other media capable of storing program verification codes. These media store computer programs, which, when executed by a processor, can implement the various steps of the aforementioned seismic signal denoising method based on locally coherent constraint-improved Radon transform.

[0038] Example 3: This embodiment proposes a computer program product, including a computer program or instructions, which, when executed by a processor, implements the seismic signal denoising method based on the improved Radon transform with local coherence constraints.

[0039] Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a computer program product.

[0040] The various embodiments in this application 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.

[0041] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the present invention.

Claims

1. A seismic signal denoising method based on locally coherent constraint-improved Radon transform, characterized in that, Includes the following steps: Read raw, noisy seismic signal data; Performing a Radon transform on the original noisy seismic signal yields the noisy seismic signal. -p field representation; based on The propagation slowness parameter p of different seismic events in the -p domain is used to perform time correction and alignment processing on the original seismic signal along the corresponding slowness. A sliding window is set along the spatial direction on the aligned seismic signal data, and the local similarity coefficient is calculated at each location, and the maximum local similarity coefficient is extracted. Using the maximum local similarity coefficient as a local coherence constraint factor The Radon transform values ​​within the -p domain are then soft-weighted to obtain the weighted result. - Values ​​of each Radon transform within the p domain; For weighted The -p domain Grardon transform is used to perform an inverse transform to obtain the denoised and reconstructed seismic signal.

2. The seismic signal denoising method based on locally coherent constraint-improved Radon transform according to claim 1, characterized in that, The original noisy seismic signal data is represented as follows: ,in, This indicates the trace number of each receiver channel in the seismic data or the spatial sampling location of the seismic record. Indicates the time sampling point.

3. A seismic signal denoising method based on locally coherent constraint-improved Radon transform according to claim 2, characterized in that, based on The propagation slowness parameter p for different seismic events in the -p domain is used to perform time correction and alignment processing on the original seismic signal along the corresponding slowness, so that seismic events that meet the slowness condition are aligned in the spatial direction. Its expression is: ； in, To adjust the original seismic signal according to the time correction relationship t= p, the original seismic signal is adjusted for the slowness parameter p of the seismic event propagation. The signal obtained after alignment is used to characterize the spatial consistency of the data in each channel under this propagation slowness condition. This is the intercept time.

4. A seismic signal denoising method based on locally coherent constraint-improved Radon transform according to claim 3, characterized in that, The formula for calculating the local similarity coefficient is as follows: ； in, For the first Within a sliding window -p domain Local similarity coefficient at location, Indicates the first A sliding window in space, This indicates the number of valid sample points participating in the calculation within the current window. As a stabilizing factor; During the movement of the sliding window, the same The maximum local similarity coefficient is obtained by taking the maximum value of the local similarity coefficients calculated from different windows at a given location. The calculation formula is as follows: 。 5. A seismic signal denoising method based on locally coherent constraint-improved Radon transform according to claim 4, characterized in that, The weighted The formulas for calculating the values ​​of each Radon transform within the -p domain are shown below: ； in, The original seismic signal is transformed by Radon transform and mapped to a coordinate system with slowness and intercept time. -p domain The value of the Radon transform at the location, For weighted The values ​​of each Radon transform within the -p domain The weighting factor is calculated using the following formula: ； In the formula, This represents the lower limit of the local similarity coefficient. This is the coherence weighting index.

6. A seismic signal denoising method based on locally coherent constraint-improved Radon transform according to claim 5, characterized in that, After soft weighting The inverse transform of the -p domain Radon transform yields the denoised and reconstructed seismic signal, as shown in the following formula: 。 7. A seismic signal denoising method based on locally coherent constraint-improved Radon transform according to claim 6, characterized in that, When the maximum local similarity coefficient Less than the preset threshold At that time, for the corresponding Additional attenuation constraints are applied at the -p domain location to further reduce the energy of low-coherence background noise.

8. A seismic signal denoising method based on locally coherent constraint-improved Radon transform according to claim 7, characterized in that, The method employs a multiplicative attenuation method for the corresponding... - Additional attenuation constraints are applied to the p-domain location.

9. A computer-readable storage medium storing executable instructions for performing the seismic signal denoising method based on locally coherent constraint-improved Radon transform as described in any one of claims 1-7, characterized in that, When the instruction is executed, it causes the processor to perform the seismic signal denoising method based on the improved Radon transform with local coherence constraints.

10. A computer program product for executing the seismic signal denoising method based on locally coherent constraint improved Radon transform as described in any one of claims 1-7, characterized in that, This includes a computer program or instructions that, when executed by a processor, implement the described seismic signal denoising method based on the improved Radon transform with local coherence constraints.

Images