Iterative method guided filtering random noise suppression method
Through the iterative method, the filtering method is guided by the filtering method, and iterative calculation and filtering is used to perform iterative calculation and filtering, which solves the problem of random noise suppression in pre-stack and post-stack seismic data, improves the signal-to-noise ratio and data fidelity, and provides more effective processing support.
Patent Information
- Application Number
- CN202310543915.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-15
- Publication Date
- 2024-11-15
AI Technical Summary
The prior art is difficult to effectively suppress random noise in seismic data processing, especially in pre-stack and post-stack seismic data, with low signal-to-noise, affecting the processing effect of the data.
The iterative guide filtering method is used to use the seismic data itself as the guide image. By calculating the mean, correlation coefficient, variance and covariance, calculating linear correlation parameters, iterative guide filtering is performed, and the signal-to-noise ratio is gradually improved and random noise is suppressed.
It improves the signal-to-noise ratio of seismic data, enhances the noise suppression accuracy, maintains the imaging effect of geological characteristics, and improves the fidelity and processing efficiency of data.
Smart Images

Figure CN118962805A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of oilfield seismic data processing, in particular to an iterative method for guiding filtering and suppressing random noise. Background Art
[0002] Guided filtering was proposed by He Kaiming et al. in 2010. This method has good edge preservation characteristics and does not experience gradient reversal. It can be widely used in image denoising, dehazing, high dynamic range compression and other fields.
[0003] In the field of seismic data processing, guided filtering is rarely used, mainly due to the lack of data with high signal-to-noise ratio as guiding images.
[0004] Chinese patent application number CN201310522885.7 relates to the extraction of a full-waveform inversion noise filter operator and its use in noise filtering. This invention, after obtaining the full-waveform inversion gradient for a single shot during full-waveform inversion, performs a two-dimensional Fourier transform on the two-dimensional gradient to convert the data from the spatial domain to the wavenumber domain. The relationship between vertical and horizontal wavenumbers in the wavenumber domain is established based on the possible range of full-waveform inversion velocity noise and the distribution of effective full-waveform inversion velocity in the wavenumber domain. A full-waveform inversion noise filter operator is employed to avoid the Gibbs effect caused by the conversion of data between the spatial and wavenumber domains. Finally, a two-dimensional inverse Fourier transform is used to obtain full-waveform inversion results with a high signal-to-noise ratio and high fidelity that are adapted to geological characteristics. This invention can determine the range of full-waveform inversion noise based on geological structural characteristics, eliminating inversion noise while maintaining effective full-waveform inversion velocity, achieving stable and reliable results.
[0005] In the Chinese patent application with application number: CN202011094102.6, a method and system for suppressing black triangle noise of a controllable vibrator is involved, which belongs to the field of data processing of seismic data excited by a controllable vibrator. The method for suppressing black triangle noise of a controllable vibrator first performs a Fourier transform on the seismic data, then uses the correlation method to solve the noise suppression operator matrix in different frequency bands, and finally uses a sparse constrained inversion iterative algorithm to achieve effective suppression of the black triangle noise of the controllable vibrator. The invention adopts a sparse constrained inversion iterative method to suppress the "black triangle" noise of the controllable vibrator, and achieves the minimum difference through difference matching and constrained inversion. Through repeated iterative updates, the optimal noise suppression processing can be achieved, and the data adaptability of the invention is better.
[0006] A Chinese patent application, filed with application number CN201410416113.X, describes a method and system for suppressing source coding crosstalk noise in full-waveform inversion. The method includes low-pass filtering the updated velocity model during the inversion iteration; performing principal component analysis on the low-pass filtered velocity model to determine the dominant signal direction of the geological structure; and performing directional filtering along the dominant signal direction to suppress source coding crosstalk noise that differs from the dominant signal direction. This invention utilizes the directional differences between the signal and noise in the full-waveform inversion velocity model to filter in the direction of the valid signal, thereby protecting the integrity of the valid signal while removing noise.
[0007] In the Chinese patent application with application number CN201210109770.0, a multiple wave suppression method is disclosed, which belongs to the field of oil, gas and coalbed methane seismic exploration and development. The method of the present invention comprises: (1) selecting a seismic trace; (2) intercepting data; (3) determining the period of the multiple wave; (4) defining the target multiple wave time window; (5) calculating the target multiple wave to obtain a filter operator; (6) convolution and filtering; (7) autocorrelation detection of the filter result; (8) iterative processing; (9) taking the seismic trace data of all other seismic traces in the original seismic data as the selected seismic trace data x(t), and repeating steps (4) to (8) for each selected seismic trace data x(t). The method of the present invention adjusts the value of the attenuation coefficient for each multiple wave, obtains a satisfactory filtering effect, and effectively suppresses the multiple waves; the method of the present invention is applicable to two-dimensional seismic data and three-dimensional seismic data.
[0008] The above existing technologies are all significantly different from the present invention and fail to solve the technical problem we want to solve. Therefore, we have invented a new iterative method to guide filtering and suppress random noise. Summary of the Invention
[0009] The purpose of the present invention is to provide an iterative method for guiding filtering and suppressing random noise, which can be applied to pre-stack and post-stack seismic data and improves the signal-to-noise ratio of seismic data.
[0010] The object of the present invention can be achieved by the following technical measures: an iterative method guided filtering method for suppressing random noise, the iterative method guided filtering method for suppressing random noise comprising:
[0011] Step 1, input seismic data as input image;
[0012] Step 2: Apply guided filtering, using the seismic data itself as the guide image, and the denoised data as the guide image in the future, to perform guided filtering to improve the signal-to-noise ratio of the data;
[0013] Step 3, suppress the random noise;
[0014] Step 4: Output the seismic data after random noise suppression.
[0015] The purpose of the present invention can also be achieved by the following technical measures:
[0016] In step 2, guided filtering is applied, and the seismic data p itself is used as the guided image I. In the future, the denoised data is used as the guided image I to perform guided filtering on the seismic data p to improve the signal-to-noise ratio of the data.
[0017] Step 2 includes:
[0018] 21) Calculate the mean of the guide image I I , mean of earthquake data p p , the correlation coefficient corr of the guided image I I , the correlation coefficient corr between the guided image I and the seismic data p Ip ;
[0019] 22) Calculate the variance var of the guide image I I , the covariance cov between the guided image I and the seismic data p Ip ;
[0020] 23) Calculate linear correlation parameters a and b;
[0021] 24) Calculate the mean of a a , the mean of b b ;
[0022] 25) Calculate the guided filtering result q.
[0023] In step 21, the mean of the guide image I is calculated using the following formula: I , mean of earthquake data p p , the correlation coefficient corr of the guided image I I , the correlation coefficient corr between the guided image I and the seismic data p Ip ;
[0024] mean I =f mean (I))
[0025] mean p =f mean (p)
[0026] corr I =f mean (I.*I)
[0027] corr Ip =f mean (I.*p)
[0028] fmean is the arithmetic mean. * indicates matrix multiplication.
[0029] Step 22: Calculate the variance var of the guide image I using the following formula: I , the covariance cov between the guided image I and the seismic data p Ip ;
[0030] var I =corr I -mean I. *mean I
[0031] cov Ip =corr Ip -mean I .*meam p .
[0032] In step 23, the linear correlation parameters a and b are calculated using the following formulas:
[0033] a=cov Ip . / (var I +∈)
[0034] b=mean p -a.*mean I
[0035] a and b are linear correlation parameters, var I is the variance of the guide image I, cv Ip is the covariance between the guided image I and the seismic data p, mean I is the mean of the guide image I, mean p is the mean of the seismic data p, and ∈ is the threshold for defining the smooth area and the edge area.
[0036] In step 24, use the following formula to calculate the mean of a a , the mean of b b ;
[0037] mean a =f mean (a)
[0038] mean b =f mean (b).
[0039] In step 25, the guided filtering result q is calculated using the following formula:
[0040] q=mean a .*I+mean b .
[0041] In step 3, the denoised data is used as the input image p, and step 2 is repeated until the number of iterations reaches the preset value and the random noise is well suppressed.
[0042] In step 3, the preset value is the seismic data itself.
[0043] The iterative guided filtering method for suppressing random noise in the present invention adopts the concept of iterative guided filtering to achieve random noise suppression in seismic data, improve noise suppression accuracy, and enhance processing efficiency. The method of the present invention can be applied to both pre-stack and post-stack seismic data, improving the signal-to-noise ratio of seismic data and providing effective technical support and guarantee for subsequent processing and interpretation. Compared with existing seismic data correction technologies, the present invention has two main advantages:
[0044] (1) Using the edge-preserving characteristics of guided filtering, the imaging effect of special geological bodies such as faults can be maintained while removing noise;
[0045] (2) In the denoising process, the concept of iterative denoising is introduced. While denoising, the effective signal is retained as much as possible to improve the fidelity and amplitude preservation of the data. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 A flowchart of a specific embodiment of the method for suppressing random noise by iterative filtering guided by the present invention;
[0047] Figure 2 is a schematic diagram of shot gather data before random noise suppression in one embodiment of the present invention;
[0048] Figure 3 Schematic diagram of shot gather data after the first round of guided filtering in one embodiment of the present invention;
[0049] Figure 4 Schematic diagram of shot gather data after the last round of guided filtering in one embodiment of the present invention;
[0050] Figure 5 A schematic diagram of a cross section before random noise suppression in one embodiment of the present invention;
[0051] Figure 6 Schematic diagram of the cross section after the last round of guided filtering in one embodiment of the present invention. DETAILED DESCRIPTION
[0052] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.
[0053] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations and / or combinations thereof.
[0054] The iterative method guided filtering method for suppressing random noise of the present invention comprises:
[0055] Step 1, using input seismic data as input image;
[0056] Step 2: Apply guided filtering, using the seismic data itself as the guide image, and the denoised data as the guide image in the future, to perform guided filtering to improve the signal-to-noise ratio of the data;
[0057] Step 3, suppress the random noise;
[0058] Step 4: Output the seismic data after random noise suppression.
[0059] The following are several specific embodiments of the present invention:
[0060] Example 1
[0061] like Figure 1 As shown, Figure 1 The flowchart of the method for suppressing random noise by iterative method guided filtering of the present invention is as follows. The method for suppressing random noise by iterative method guided filtering includes:
[0062] Step 101, using input seismic data as input image p;
[0063] Step 102 is to apply guided filtering for the first time, using the seismic data p itself as the guided image I. Subsequently, the denoised data is used as the guided image I to perform guided filtering on the seismic data p to improve the signal-to-noise ratio. The guided filtering process is divided into the following five steps:
[0064] 1) Use the following formula to calculate the mean of the guide image I I , mean of earthquake data p p , the correlation coefficient corr of the guided image I I , the correlation coefficient corrI between the guided image I and the seismic data p p ;
[0065] mean I =f mean I)
[0066] mean p =fmean (p)
[0067] corr I =f mean (I.*I)
[0068] corr Ip =f mean (I.*p)
[0069] 2) Use the following formula to calculate the variance var of the guide image I I , the covariance cov between the guided image I and the seismic data p Ip ;
[0070] var I =corr I -mean I .*mean I
[0071] cov Ip =corr Ip -mean I .*mean p
[0072] 3) Use the following formulas to calculate the linear correlation parameters a and b respectively;
[0073] a=cov Ip . / (var I +∈)
[0074] b=mean p -a.*mean I
[0075] 4) Use the following formula to calculate the mean of a a , the mean of b b ;
[0076] mean a =f mean (a)
[0077] mean h =f mean (b)
[0078] 5) Calculate the guided filtering result q using the following formula;
[0079] q=mean a .*I+mean b
[0080] Step 103: Using the denoised data as the input image p, repeat step 2 until the number of iterations reaches a preset value and the random noise is well suppressed.
[0081] Step 104: output the seismic data after random noise is suppressed.
[0082] Example 2
[0083] In a specific embodiment 2 of the present invention, the method for suppressing random noise by iteratively guided filtering includes the following steps:
[0084] (1) Figure 2 As shown, the input seismic data is used as the input image;
[0085] (2) Guided filtering is applied for the first time, using the seismic data itself as the guide image, and then using the denoised data as the guide image to perform guided filtering to improve the signal-to-noise ratio of the data, such as Figure 3 As shown;
[0086] (3) Using the denoised data as the input image, repeat step (2) until the number of iterations reaches the preset value and the random noise is well suppressed, such as Figure 4 As shown;
[0087] (4) Output seismic data after random noise suppression.
[0088] Example 3
[0089] In a specific embodiment 3 of the present invention, for stacked sections with a low signal-to-noise ratio, the iterative method for filtering and suppressing random noise in the sections includes the following steps:
[0090] (1) Take the input seismic profile as the input image, such as Figure 5 ;
[0091] (2) Guided filtering is applied for the first time, with the profile itself as the guide image. In the following cases, the profile after the last denoising is used as the guide image for guided filtering to improve the signal-to-noise ratio of the data.
[0092] (3) Using the denoised cross-section as the input image, repeat step (2) until the number of iterations reaches the preset value and the random noise is well suppressed;
[0093] (4) Output the seismic profile after random noise suppression, such as Figure 6 .
[0094] The iterative guided filtering method of this invention employs the concept of iterative guided filtering to suppress random noise in seismic data, improving noise suppression accuracy and processing efficiency. The method can be applied to both pre-stack and post-stack seismic data, improving the signal-to-noise ratio of the data and providing effective technical support and assurance for subsequent processing and interpretation.
[0095] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art may modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features therein. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.
[0096] Except for the technical features described in the specification, all other technical features are known technologies to those skilled in the art.
Claims
1. Iterative method guided filtering method to suppress random noise, characterized in that: The iterative method guides the filtering to suppress random noise, including: Step 1, input seismic data as input image; Step 2: Apply guided filtering, use the seismic data itself as the guided image, and then use the denoised data as the guided image to perform guided filtering to improve the signal-to-noise ratio of the data; Step 3, suppressing random noise; Step 4: Output the seismic data after random noise suppression.
2. The method for suppressing random noise by iterative filtering according to claim 1, characterized in that: In step 2, guided filtering is applied, and the seismic data p itself is used as the guided image I. In the future, the denoised data is used as the guided image I to perform guided filtering on the seismic data p to improve the signal-to-noise ratio of the data.
3. The method for suppressing random noise by iterative filtering according to claim 2, characterized in that: Step 2 includes: 21) Calculate the mean of the guide image I I , mean of earthquake data p p , the correlation coefficient of the guided image I is corr I , the correlation coefficient corr between the guided image I and the seismic data p Ip ; 22) Calculate the variance var of the guidance image I I , the covariance cov between the guided image I and the seismic data p Ip ; 23) Calculate linear correlation parameters a and b; 24) Calculate the mean of a a , the mean of b b ; 25) Calculate the guided filtering result q.
4. The method for suppressing random noise by iterative filtering according to claim 3, characterized in that: In step 21, the mean of the guide image I is calculated using the following formula: I , mean of earthquake data p p , the correlation coefficient of the guided image I is corr I , the correlation coefficient corr between the guided image I and the seismic data p Ip ; mean I =f mean (I) mean p =f mean (p) corr I =f mean (I.*I) corr Ip =f mean (I.*p) fmean is the arithmetic mean. * indicates matrix multiplication.
5. The method for suppressing random noise by iterative filtering according to claim 4, characterized in that: In step 22, the variance var of the guide image I is calculated using the following formulas: I , the covariance cov between the guided image I and the seismic data p Ip ; var I =odd I -mean I .*mean I coV Ip =corr Ip -mean I ·*mean p 。 6. The method for suppressing random noise by iterative filtering according to claim 5, characterized in that: In step 23, the linear correlation parameters a and b are calculated respectively using the following formulas; a=cov Ip · / (var I +∈) b=mean p -a.*mean I a and b are linearly related parameters, var I is the variance of the guide image I, cov Ip is the covariance of the guided image I and the seismic data p, mean I is the mean of the guide image I, mean p is the mean of the seismic data p, and ∈ is the threshold for defining the smooth area and the edge area.
7. The method for suppressing random noise by iteratively guided filtering according to claim 6, characterized in that: In step 24, use the following formula to calculate the mean of a a , the mean of b b ; mean a =f mean (a) mean b =f mean (b)。 8. The method for suppressing random noise by iteratively guided filtering according to claim 7, characterized in that: In step 25, the guided filtering result q is calculated using the following formula; q=mean a .*I+mean b 。 9. The method for suppressing random noise by iterative filtering according to claim 1, characterized in that: In step 3, the denoised data is used as the input image p, and step 2 is repeated until the number of iterations reaches a preset value and the random noise is well suppressed.
10. The method for suppressing random noise by iteratively guided filtering according to claim 9, characterized in that: In step 3, the preset value is the seismic data itself.
Citation Information
Patent Citations
Multiple suppression method
CN103376466A
Method for extracting full-waveform inversion noise filtering operator and performing noise filtering through full-waveform inversion noise filtering operator
CN104570090A
Method and system for suppression of source coding crosstalk noise of full waveform inversion
CN105353409A
Vibroseis black triangle noise suppression method and system
CN114428345A
