技术领域technical field
本发明涉及地震勘探领域，尤其涉及油气地球物理勘探中的地震资料处理领域，具体的讲是一种在地震勘探中，对地震数据随机噪声进行衰减的方法及系统。The invention relates to the field of seismic exploration, in particular to the field of seismic data processing in oil and gas geophysical exploration, specifically a method and system for attenuating random noise of seismic data in seismic exploration.
背景技术Background technique
地震勘探是一种利用人工地震技术探测地下结构的勘探方法。它按照一定的方式人工激发地震波，利用称之为检波器的装置接收来自地下的反射信号，通过对反射信号的处理和分析探测地下结构。Seismic exploration is an exploration method that uses artificial seismic technology to detect underground structures. It artificially stimulates seismic waves in a certain way, uses a device called a geophone to receive reflection signals from the ground, and detects underground structures through processing and analysis of reflection signals.
但是，检波器在接收地震信号的同时，也接收到大量来自地下和地表的随机噪声，降低了地震记录信噪比，严重干扰地震信号反映地下结构的能力，随机噪声的衰减与弱信号恢复是地震资料处理领域的重要研究内容。However, while the geophone receives seismic signals, it also receives a large amount of random noise from the underground and the surface, which reduces the signal-to-noise ratio of seismic records and seriously interferes with the ability of seismic signals to reflect underground structures. The attenuation of random noise and the recovery of weak signals are important Important research content in the field of seismic data processing.
1、f-x域空间预测滤波是目前工业界应用最为广泛、效果最为稳定的随机噪声衰减方法。该方法基于信号在频率－空间域的可预测性和随机噪声的不可预测性进行信号识别和噪声压制。其滤波算子的求取已经由自回归模型(AR)因果算子发展到非因果算子(Gulunay,2000)，去噪算法也由二维发展到三维甚至四维去噪。但是，该方法要求地震信号在横向上具有稳态特征，且信号在横向上呈局部线性趋势。对于横向非均质性较强、构造较为复杂的地层结构而言，其反射特征很难满足f-x域空间预测滤波的上述要求，不能取得理想的去噪效果。1. The f-x domain spatial prediction filter is currently the most widely used and most stable random noise attenuation method in the industry. The method performs signal identification and noise suppression based on the predictability of the signal in the frequency-space domain and the unpredictability of the random noise. The calculation of the filter operator has been developed from the autoregressive model (AR) causal operator to the non-causal operator (Gulunay, 2000), and the denoising algorithm has also been developed from two-dimensional to three-dimensional or even four-dimensional denoising. However, this method requires that the seismic signal has a steady-state characteristic in the lateral direction, and the signal has a local linear trend in the lateral direction. For stratigraphic structures with strong lateral heterogeneity and complex structures, its reflection characteristics are difficult to meet the above requirements of f-x domain spatial prediction filtering, and ideal denoising effects cannot be achieved.
2、时频分析类方法可以针对非稳态以及非线性信号进行处理，该类方法是近年来随着小波变换等时频分析工具的兴起而发展起来。根据有效信号与随机噪声在时频域的分布差异，首先通过小波变换等时频分析工具将地震数据变换到时频域，再选择合适的时频域滤波手段将有效信号与噪声分离，然后反变换到时间域得到去噪后的结果。但是，该类方法的数学变换在实际地震资料处理中缺乏明确的物理意义，没有考虑地震信号本身的固有特点，且实现过程复杂，可操作性较差，制约了该方法由实验室到工业界应用的转换。2. Time-frequency analysis methods can deal with unsteady and nonlinear signals. This type of method has been developed with the rise of time-frequency analysis tools such as wavelet transform in recent years. According to the distribution difference between the effective signal and random noise in the time-frequency domain, the seismic data is first transformed into the time-frequency domain through time-frequency analysis tools such as wavelet transform, and then the appropriate time-frequency domain filtering method is selected to separate the effective signal from the noise. Transform to the time domain to get the denoised result. However, the mathematical transformation of this type of method lacks clear physical meaning in the actual seismic data processing, does not consider the inherent characteristics of the seismic signal itself, and the implementation process is complicated and the operability is poor, which restricts the method from the laboratory to the industry. Applied transformations.
3、希尔伯特黄变换（HHT）是一种基于信号本身固有属性的时频分析方法，该方法将非平稳信号通过经验模态分解（EMD）方式分解为不同尺度的平稳窄带信号，称为固有模态函数(IMF)，然后对这些固有模态函数进行希尔伯特变换就得到信号的时频谱。该方法克服了小波变换需要选取固定小波基的弱点，所分解的固有模态函数反映了信号本身的固有属性，具有明确的物理意义，更加有效的反映了地震信号的多尺度特征。Ivan（1999）第一次将经验模态分解方法引入到地震资料处理领域，主要用于地震属性分析和提高分辨率处理。Bekara（2008）将经验模态分解与f-x滤波相结合，提出了一种新的f-x域随机噪声衰减方法，首次将经验模态分解技术引入到地震数据随机噪声衰减方法的研究。该方法利用经验模态分解代替常规f-x域预测滤波的线性自回归滤波器，通过对频率切片进行经验模态分解并去除第一个固有模态函数，达到噪声压制的目的。虽然随机噪声经过经验模态分解之后，大部分能量集中在第一个模态分量上，但其它模态分量上依然有较强的残存能量，且复杂反射的有效信号也会泄露在第一模态分量上，因此，仅仅通过剔除某一分量或某些分量的滤波方法在去噪能力和保幅性能上存在较大缺陷。3. Hilbert-Huang transform (HHT) is a time-frequency analysis method based on the inherent properties of the signal itself. This method decomposes non-stationary signals into stationary narrow-band signals of different scales through empirical mode decomposition (EMD), called is the intrinsic mode function (IMF), and then Hilbert transforms these intrinsic mode functions to obtain the time spectrum of the signal. This method overcomes the weakness that wavelet transform needs to select a fixed wavelet basis. The decomposed intrinsic mode function reflects the inherent properties of the signal itself, has clear physical meaning, and reflects the multi-scale characteristics of seismic signals more effectively. Ivan (1999) introduced the empirical mode decomposition method into the field of seismic data processing for the first time, mainly for seismic attribute analysis and resolution improvement processing. Bekara (2008) combined empirical mode decomposition and f-x filtering, and proposed a new f-x domain random noise attenuation method, which was the first time to introduce empirical mode decomposition technology into the study of seismic data random noise attenuation method. In this method, empirical mode decomposition is used to replace the linear autoregressive filter of conventional f-x domain predictive filtering, and the purpose of noise suppression is achieved by performing empirical mode decomposition on frequency slices and removing the first intrinsic mode function. Although most of the energy of random noise is concentrated in the first mode component after empirical mode decomposition, there is still strong residual energy in other mode components, and the effective signal of complex reflection will also leak in the first mode component. Therefore, the filtering method that only eliminates a certain component or certain components has relatively large defects in denoising ability and amplitude preservation performance.
发明内容Contents of the invention
本发明的目的是为了克服现有技术中存在的衰减随机噪声信号不够理想的不足，提供一种衰减地震数据随机噪声的方法与系统，以解决上述问题。The purpose of the present invention is to overcome the disadvantages of attenuating random noise signals in the prior art that are not ideal, and provide a method and system for attenuating random noise in seismic data to solve the above problems.
为了达到上述目的，本发明实施例公开了一种衰减地震数据随机噪声的方法，包括：获取地震数据；将所述地震数据进行傅立叶变换，生成频率-空间域的地震数据；在空间方向上对所述频率-空间域的地震数据进行复数经验模态分解，生成多个模态分量；根据所述频率-空间域的地震数据与所述多个模态分量，利用最优化方法生成自适应信号重构算子；根据所述自适应信号重构算子和所述多个模态分量，重构生成频率域地震信号；将所述频率域地震信号进行傅立叶反变换，生成随机噪声衰减后的时间域地震信号；根据所述随机噪声衰减后的时间域地震信号绘制噪声衰减后的地震剖面图像。In order to achieve the above purpose, the embodiment of the present invention discloses a method for attenuating random noise in seismic data, including: acquiring seismic data; performing Fourier transform on the seismic data to generate seismic data in the frequency-space domain; The seismic data in the frequency-space domain is subjected to complex empirical mode decomposition to generate multiple modal components; according to the seismic data in the frequency-space domain and the multiple modal components, an adaptive signal is generated using an optimization method A reconstruction operator; according to the adaptive signal reconstruction operator and the plurality of modal components, reconstruct and generate a frequency domain seismic signal; perform an inverse Fourier transform on the frequency domain seismic signal to generate a random noise attenuated A time-domain seismic signal; drawing a noise-attenuated seismic profile image according to the time-domain seismic signal after the random noise attenuation.
为了达到上述目的，本发明实施例还公开了一种衰减地震数据随机噪声的系统，包括：地震数据获取单元，用于获取地震数据；频率-空间域的地震数据生成单元，用于将所述地震数据进行傅立叶变换，生成频率-空间域的地震数据；模态分量生成单元，用于在空间方向上对所述频率-空间域的地震数据进行复数经验模态分解，生成多个模态分量；重构算子生成单元，用于根据所述频率-空间域的地震数据与所述多个模态分量，利用最优化方法生成自适应信号重构算子；频率域地震信号生成单元，用于根据所述自适应信号重构算子和所述多个模态分量，重构生成频率域地震信号；时间域地震信号生成单元，用于将所述频率域地震信号进行傅立叶反变换，生成随机噪声衰减后的时间域地震信号；地震剖面图像绘制单元，用于根据所述随机噪声衰减后的时间域地震信号绘制噪声衰减后的地震剖面图像。In order to achieve the above purpose, the embodiment of the present invention also discloses a system for attenuating random noise of seismic data, including: a seismic data acquisition unit for acquiring seismic data; a frequency-space domain seismic data generation unit for converting the The seismic data is subjected to Fourier transform to generate seismic data in the frequency-space domain; the modal component generation unit is used to perform complex empirical mode decomposition on the seismic data in the frequency-space domain in the spatial direction to generate multiple modal components The reconstruction operator generation unit is used to generate an adaptive signal reconstruction operator using an optimization method according to the seismic data in the frequency-space domain and the plurality of modal components; the frequency domain seismic signal generation unit uses According to the adaptive signal reconstruction operator and the plurality of modal components, the frequency domain seismic signal is reconstructed and generated; the time domain seismic signal generating unit is used to perform inverse Fourier transform on the frequency domain seismic signal to generate The time-domain seismic signal after random noise attenuation; the seismic section image drawing unit, configured to draw a noise-attenuated seismic section image according to the time-domain seismic signal after random noise attenuation.
本发明实施例的衰减地震数据随机噪声的方法与系统，可以自动识别并重构被随机噪声严重污染的地震信号，有效压制随机噪声对地震信号的影响，提高了地震资料信噪比，增强了地震信号对地下复杂结构和油气储层的检测精度，为地震资料构造解释和储层预测提供了高质量的基础数据。The method and system for attenuating random noise in seismic data according to the embodiments of the present invention can automatically identify and reconstruct seismic signals seriously polluted by random noise, effectively suppress the influence of random noise on seismic signals, improve the signal-to-noise ratio of seismic data, and enhance the The detection accuracy of seismic signals for complex underground structures and oil and gas reservoirs provides high-quality basic data for structural interpretation of seismic data and reservoir prediction.
附图说明Description of drawings
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained according to these drawings on the premise of not paying creative efforts.
图1为本发明实施例中的复数经验模态分解介绍中的高斯白噪声EMD分解后的各IMF频谱示意图；Fig. 1 is each IMF frequency spectrum schematic diagram after Gaussian white noise EMD decomposition in the introduction of complex empirical mode decomposition in the embodiment of the present invention;
图2为利用投影法进行复数经验模态分解的示意图；Fig. 2 is the schematic diagram that utilizes projection method to carry out complex empirical mode decomposition;
图3为本发明实施例的衰减地震数据随机噪声的方法的流程图；3 is a flowchart of a method for attenuating random noise in seismic data according to an embodiment of the present invention;
图4为本发明实施例的衰减地震数据随机噪声的系统的结构示意图；4 is a schematic structural diagram of a system for attenuating random noise in seismic data according to an embodiment of the present invention;
图5为本发明实施例1中采集的某油田A区块的碳酸盐岩地震记录；Fig. 5 is the carbonate rock seismic record of certain oilfield A block collected in the embodiment of the present invention 1;
图6为本发明实施例1中的经过f-x域空间预测滤波之后的某油田A区块碳酸盐岩地震记录；Fig. 6 is the carbonate rock seismic record of block A of an oil field after f-x domain spatial prediction and filtering in Embodiment 1 of the present invention;
图7为本发明实施例1中的经过本发明的衰减方法处理之后的某油田A区块碳酸盐岩地震记录；Fig. 7 is the carbonate rock seismic record of block A of an oil field after being processed by the attenuation method of the present invention in Embodiment 1 of the present invention;
图8为本发明实施例2中采集的某油田B区块的地震记录；Fig. 8 is the seismic record of a block B of an oilfield collected in Embodiment 2 of the present invention;
图9为本发明实施例2中的经过本发明的衰减方法处理之后的某油田B区块的地震记录。Fig. 9 is the seismic record of block B of an oil field after being processed by the attenuation method of the present invention in Example 2 of the present invention.
具体实施方式Detailed ways
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.
在介绍本发明的具体实施方式之前，为了更好地理解本发明的实施过程，首先对本发明所涉及的复数经验模态分解进行简单的介绍。Before introducing the specific implementation of the present invention, in order to better understand the implementation process of the present invention, a brief introduction to the complex empirical mode decomposition involved in the present invention is firstly made.
N E.Huang等人认为，平稳信号应该满足两个条件：（1）极值点数目和过零点数目相等或最多相差1个；（2）在任意点，由局部极大值点和局部极小值点构成的两条包络线平均值为0。满足该条件的信号称为固有模态信号，相应的函数称为固有模态函数(Intrinsic Mode Function,简记IMF)。对于复杂非平稳信号，并不满足IMF条件，因此N E.Huang等人提出了如下假设：任何信号都是由一些不同的固有模态组成的；每个模态可以是线性的，也可以是非线性的，其极点数和零点数相同，且上下包络线关于时间轴局部对称；任何时候，一个信号都可以包含许多固有模态信号；如果模态之间相互重叠，便形成复合信号。将复合信号分解为平稳窄带固有模态信号的过程称为经验模态分解(Empirical Mode Decomposition,EMD），其目的是将复杂非平稳信号转化为平稳信号，为后续处理带来方便。其核心就是对信号进行不断“筛分”，首先找出原始数列的局部极大值和极小值，利用三次样条插值连接局部极大值和极小值，分别得到极大值包络和极小值包络，然后对每个时刻的局部极大值和极小值取平均，得到瞬时平均值m(t)：N E. Huang et al. believe that a stationary signal should satisfy two conditions: (1) the number of extreme points and the number of zero-crossing points are equal or differ by at most 1; (2) at any point, the local maximum point and the local extreme point The average value of the two envelopes formed by the small value points is 0. The signal that satisfies this condition is called an intrinsic mode signal, and the corresponding function is called an intrinsic mode function (IMF for short). For complex non-stationary signals, the IMF condition is not satisfied, so N E.Huang et al. put forward the following hypothesis: any signal is composed of some different intrinsic modes; each mode can be linear or non-linear Linear, with the same number of poles and zeros, and the upper and lower envelopes are locally symmetrical about the time axis; at any time, a signal can contain many natural mode signals; if the modes overlap with each other, a composite signal is formed. The process of decomposing a composite signal into a stationary narrow-band intrinsic mode signal is called Empirical Mode Decomposition (EMD), and its purpose is to convert a complex non-stationary signal into a stationary signal for convenience in subsequent processing. Its core is to continuously "screen" the signal. First, find the local maximum and minimum values of the original sequence, and use cubic spline interpolation to connect the local maximum and minimum values to obtain the maximum envelope and The minimum envelope, and then average the local maximum and minimum values at each moment to obtain the instantaneous average value m(t):
考虑原始数列减去瞬时平均值得到的新数列是否满足固有模态函数两个条件，若满足，则将其作为一个固有模态函数，若不满足，则作为原始数列重复上述过程，直到满足条件为止。这样，得到第一个固有模态函数c1(t)，将其从原始数列中分离出来，然后将余数r1(t)作为新的序列进行以上分解过程，直到剩余项rn(t)变为单调函数或者常数，此时无IMF分解出来，从而结束分解，得到从原始数列中分离出的n个固有模态函数分量和一个趋势项rn(t)，分解后信号可表示为：Consider whether the new sequence obtained by subtracting the instantaneous average value from the original sequence satisfies the two conditions of the intrinsic mode function. If it is satisfied, it will be regarded as an intrinsic mode function. If it is not satisfied, the above process will be repeated as the original sequence until the conditions are met until. In this way, the first intrinsic mode function c1 (t) is obtained, which is separated from the original sequence, and then the remainder r1 (t) is used as a new sequence to carry out the above decomposition process until the remaining item rn (t) becomes a monotonic function or a constant, and no IMF is decomposed at this time, thus ending the decomposition, and obtaining n intrinsic mode function components and a trend item rn (t) separated from the original sequence, the decomposed signal can be expressed as:
由于EMD是基于信号局部特征的分解，其特征尺度参数是基于实际测量所获得的数据，因此由其分解得到的固有模态函数表征了信号在某一特征尺度上的振动模式和频率变动范围。对均匀分布的白噪信号进行EMD分解表明，任何一个IMF分量极值点数近似是前一个IMF分量极值点数的一半，其平均周期是前一个IMF分量的两倍，该特征不随数据长度的改变而改变。如图1所示，从高斯白噪声的IMF频谱图可以看出，每一个IMF主频是前一个IMF主频的一半，也就是说，EMD相当于一个二进滤波器组。这同时也说明，EMD所得到的固有模态函数与分频滤波所得到结果是不同的，固有模态函数表征了信号的快速振动分量与慢速振动分量，而且不同模态之间含有频率重叠，如果能够采用加权方式将IMF进行组合，实际上应该类似于对一定频带范围数据进行“叠加”的过程，可以有效提高信噪比。Since EMD is based on the decomposition of the local characteristics of the signal, and its characteristic scale parameters are based on the data obtained by actual measurement, the intrinsic mode function obtained by its decomposition characterizes the vibration mode and frequency range of the signal on a certain characteristic scale. The EMD decomposition of uniformly distributed white noise signals shows that the number of extremum points of any IMF component is approximately half of the number of extremum points of the previous IMF component, and its average period is twice that of the previous IMF component. This feature does not change with the length of the data. And change. As shown in Figure 1, it can be seen from the IMF spectrum diagram of Gaussian white noise that the main frequency of each IMF is half of the main frequency of the previous IMF, that is to say, EMD is equivalent to a binary filter bank. This also shows that the intrinsic mode function obtained by EMD is different from the result obtained by frequency division filtering. The intrinsic mode function characterizes the fast vibration component and slow vibration component of the signal, and there is frequency overlap between different modes. , if the IMF can be combined in a weighted manner, it should actually be similar to the process of "superimposing" data in a certain frequency range, which can effectively improve the signal-to-noise ratio.
常规的EMD方法是在针对时间域信号的分解方法，常规方法无法实现复数域信号的EMD分解。Bekara(2008)提出了对频率域复数信号的实部和虚部分别进行EMD分解，然后将分解后的结果对应组合，从而形成EMD后的复数IMF。该方法虽然解决了复数分解问题，但是其仍然是时间域的算法，实质是将复数信号映射到两个独立的实值单变量空间，两者之间存在的相互关系被扭曲了，所得到的复数分量缺乏物理意义。投影法EMD分解方式可以实现复数信号的经验模态分解，该方法首先将复数信号投影到指定方向上，找到投影向量极大值并求取极大值包络，在不同方向上重复该过程，形成一个包围信号的三维管束，管束中心定义为慢速震荡信号，将原始信号减去慢速震荡后分裂出的快速震荡，称为复数IMF，依次重复该过程得到一系列复数IMF。该方法得到的固有模态函数反映了复数信号在不同尺度上的振动情况，具有明确的物理意义。如图2为利用投影法进行复数经验模态分解，图2(a)为复数信号，图2(b)为复数信号在不同方向投影形成的管束包络，图2(c)是管束中心定义的慢速振荡分量，图2(d)是快速震荡信号，即固有模态函数。The conventional EMD method is aimed at the decomposition method of the time domain signal, and the conventional method cannot realize the EMD decomposition of the complex number domain signal. Bekara (2008) proposed to perform EMD decomposition on the real part and imaginary part of the complex signal in the frequency domain, and then combine the decomposed results correspondingly to form the complex IMF after EMD. Although this method solves the complex number decomposition problem, it is still an algorithm in the time domain. The essence is to map the complex signal to two independent real-valued univariate spaces. The relationship between the two is distorted, and the obtained Complex components lack physical meaning. The EMD decomposition method of the projection method can realize the empirical mode decomposition of the complex signal. This method first projects the complex signal to the specified direction, finds the maximum value of the projection vector and obtains the maximum value envelope, and repeats the process in different directions. A three-dimensional tube bundle surrounding the signal is formed. The center of the tube bundle is defined as the slow oscillation signal. The fast oscillation obtained by subtracting the slow oscillation from the original signal is called a complex IMF. Repeat this process in turn to obtain a series of complex IMFs. The intrinsic mode function obtained by this method reflects the vibration of the complex signal at different scales, and has a clear physical meaning. Figure 2 shows the complex empirical mode decomposition using the projection method, Figure 2(a) is the complex signal, Figure 2(b) is the envelope of the tube bundle formed by the projection of the complex signal in different directions, and Figure 2(c) is the definition of the tube bundle center The slow oscillating component of , Figure 2(d) is the fast oscillating signal, that is, the intrinsic mode function.
上面介绍了复数经验模态分解的基本原理及其模态函数的概念，本发明基于上述方法对地震数据进行复数经验模态分解，然后采用最优化方法由地震数据及其模态分量求取自适应信号重构算子，再利用自适应信号重构算子，由地震数据的不同模态分量重构噪声衰减之后的地震信号。Introduced above the basic principle of complex empirical mode decomposition and the concept of modal function thereof, the present invention carries out complex empirical mode decomposition to seismic data based on above-mentioned method, adopts optimization method to obtain from seismic data and its modal component then The adaptive signal reconstruction operator is used, and then the adaptive signal reconstruction operator is used to reconstruct the seismic signal after noise attenuation from the different modal components of the seismic data.
图3为本发明实施例的衰减地震数据随机噪声的方法的流程图，如图所示，本实施例的方法包括：Fig. 3 is the flow chart of the method for the attenuation seismic data random noise of the embodiment of the present invention, as shown in the figure, the method of the present embodiment comprises:
步骤S101，获取地震数据；步骤S102，将所述地震数据进行傅立叶变换，生成频率-空间域的地震数据；步骤S103，在空间方向上对所述频率-空间域的地震数据进行复数经验模态分解，生成多个模态分量；步骤S104，根据所述频率-空间域的地震数据与所述多个模态分量，利用最优化方法生成自适应信号重构算子；步骤S105，根据所述自适应信号重构算子和所述多个模态分量，重构生成频率域地震信号；步骤S106，将所述频率域地震信号进行傅立叶反变换，生成随机噪声衰减后的时间域地震信号；步骤S107，根据所述随机噪声衰减后的时间域地震信号绘制噪声衰减后的地震剖面图像。Step S101, acquiring seismic data; Step S102, performing Fourier transform on the seismic data to generate seismic data in the frequency-space domain; Step S103, performing complex empirical mode on the seismic data in the frequency-space domain in the spatial direction Decompose to generate multiple modal components; Step S104, according to the seismic data in the frequency-space domain and the multiple modal components, use an optimization method to generate an adaptive signal reconstruction operator; Step S105, according to the The adaptive signal reconstruction operator and the plurality of modal components are reconstructed to generate frequency-domain seismic signals; step S106, performing inverse Fourier transform on the frequency-domain seismic signals to generate time-domain seismic signals after random noise attenuation; Step S107, drawing a seismic section image after noise attenuation according to the time domain seismic signal after random noise attenuation.
在本实施例中，步骤S101获取地震数据x(i,t),i＝1,2,…n，其中n为记录道数，本实施例中，获取的地震数据的记录道数n为70，当然本发明不限于此，可以根据实际情况选择其他数值。In this embodiment, step S101 acquires seismic data x(i, t), i=1, 2,...n, where n is the number of records, and in this embodiment, the number n of seismic data acquired is 70 , of course, the present invention is not limited thereto, and other values may be selected according to actual conditions.
在本实施例中，步骤S102将所述地震数据进行傅立叶变换，生成频率-空间域的地震数据，包括：In this embodiment, step S102 performs Fourier transform on the seismic data to generate frequency-space domain seismic data, including:
对地震数据x(i,t),i＝1,2,…n进行傅立叶变换，生成频率-空间域的地震数据X(i,f),i＝1,2,…n；其中，Perform Fourier transform on seismic data x(i,t), i=1,2,...n to generate frequency-space domain seismic data X(i,f),i=1,2,...n; where,
X(i,f)=∫x(i,t)e-j2πftdt。X(i,f)=∫x(i,t)e-j2πft dt.
在本实施例中，步骤S 103在空间方向上对所述频率-空间域的地震数据进行复数经验模态分解，生成多个模态分量，包括：In this embodiment, step S103 performs complex empirical mode decomposition on the seismic data in the frequency-space domain in the spatial direction to generate a plurality of modal components, including:
对所述频率-空间域的地震数据X(i,f),i＝1,2,…n进行复数经验模态分解，生成多个模态分量Cj(i,f),j=1,2,…,m，其中，m是经过经验模态分解的模态分量的个数，在本实施例中，经过经验模态分解的模态分量的个数m为9。performing complex empirical mode decomposition on the seismic data X(i,f) in the frequency-space domain, i=1,2,...n, to generate multiple modal components Cj (i,f),j=1, 2, . . . , m, wherein, m is the number of modal components undergoing empirical mode decomposition, and in this embodiment, the number m of modal components undergoing empirical mode decomposition is 9.
在本实施例中，步骤S104根据所述频率-空间域的地震数据与所述多个模态分量，利用最优化方法生成自适应信号重构算子，包括：In this embodiment, step S104 uses an optimization method to generate an adaptive signal reconstruction operator according to the seismic data in the frequency-space domain and the multiple modal components, including:
步骤1，根据所述频率-空间域的地震数据X(i,f),i＝1,2,…n构建向量X，根据所述多个模态分量Cj(i,f),j=1,2,…,m构建矩阵C，根据所述自适应信号重构算子Aj(f),j=1,2,…,m构建向量A；Step 1, construct a vector X according to the seismic data X(i,f) in the frequency-space domain, i=1,2,...n, and according to the multiple modal components Cj (i,f),j= 1, 2,..., m constructing a matrix C, constructing a vector A according to the adaptive signal reconstruction operator Aj (f), j=1, 2,..., m;
步骤2，建立目标函数E=X-C×AT；Step 2, establish the objective function E=XC×AT ;
步骤3，利用最优化方法计算所述目标函数的最小值，生成所述目标函数最小值下的自适应信号重构算子Aj(f),j=1,2,…,m；Step 3, using an optimization method to calculate the minimum value of the objective function, and generate an adaptive signal reconstruction operator Aj (f),j=1,2,...,m under the minimum value of the objective function;
此时，A=(CCT+βI)-1CTX，其中，I为单位矩阵，β为稳定性阻尼系数，且β=0.01。At this time, A=(CCT +βI)-1 CT X, where I is the identity matrix, β is the stability damping coefficient, and β=0.01.
在本实施例中，步骤S105根据所述自适应信号重构算子和所述多个模态分量，重构生成频率域地震信号，包括：In this embodiment, step S105 reconstructs and generates frequency-domain seismic signals according to the adaptive signal reconstruction operator and the multiple modal components, including:
利用所述自适应信号重构算子Aj(f),j=1,2,…,m，由所述多个模态分量Cj(i,f),j=1,2,…,m重构所述频率域地震信号S(i,f),i＝1,2,…n：Using the adaptive signal reconstruction operator Aj (f),j=1,2,...,m, from the multiple modal components Cj (i,f),j=1,2,..., m reconstructing the frequency domain seismic signal S(i, f), i=1, 2,...n:
在本实施例中，步骤S106将所述频率域地震信号进行傅立叶反变换，生成随机噪声衰减后的时间域地震信号，包括：In this embodiment, step S106 performs an inverse Fourier transform on the frequency-domain seismic signal to generate a time-domain seismic signal after random noise attenuation, including:
对重构后得到的所述频率域地震信号进行傅立叶反变换，生成随机噪声衰减后的时间域地震信号s(i,t),i＝1,2,…n：For the frequency domain seismic signal obtained after reconstruction Perform inverse Fourier transform to generate time-domain seismic signals s(i,t) after random noise attenuation, i=1,2,...n:
s(i,t)=∫S(i,f)ej2πftdf。s(i,t)=∫S(i,f)ej2πft df.
在本实施例中，步骤S107根据常用的地震数据绘制软件就可以绘制噪声衰减后的地震剖面图像。In this embodiment, step S107 can draw the seismic section image after the noise attenuation according to the commonly used seismic data drawing software.
图4为本发明实施例的衰减地震数据随机噪声的系统的结构示意图，如图所示，本实施例的系统包括：Fig. 4 is a schematic structural diagram of a system for attenuating random noise in seismic data according to an embodiment of the present invention. As shown in the figure, the system of this embodiment includes:
地震数据获取单元101，用于获取地震数据；频率-空间域的地震数据生成单元102，用于将所述地震数据进行傅立叶变换，生成频率-空间域的地震数据；模态分量生成单元103，用于在空间方向上对所述频率-空间域的地震数据进行复数经验模态分解，生成多个模态分量；重构算子生成单元104，用于根据所述频率-空间域的地震数据与所述多个模态分量，利用最优化方法生成自适应信号重构算子；频率域地震信号生成单元105，用于根据所述自适应信号重构算子和所述多个模态分量，重构生成频率域地震信号；时间域地震信号生成单元106，用于将所述频率域地震信号进行傅立叶反变换，生成随机噪声衰减后的时间域地震信号；地震剖面图像绘制单元107，用于根据所述随机噪声衰减后的时间域地震信号绘制噪声衰减后的地震剖面图像。The seismic data acquisition unit 101 is used to acquire seismic data; the seismic data generation unit 102 in the frequency-space domain is used to perform Fourier transform on the seismic data to generate the seismic data in the frequency-space domain; the modal component generation unit 103, It is used to perform complex empirical mode decomposition on the seismic data in the frequency-space domain in the spatial direction to generate a plurality of modal components; the reconstruction operator generation unit 104 is used to and the plurality of modal components, using an optimization method to generate an adaptive signal reconstruction operator; a frequency domain seismic signal generating unit 105, configured to reconstruct the operator and the plurality of modal components according to the adaptive signal , reconstruct and generate frequency-domain seismic signals; time-domain seismic signal generation unit 106 is used to perform inverse Fourier transform on the frequency-domain seismic signals to generate time-domain seismic signals after random noise attenuation; seismic section image drawing unit 107, use A seismic section image after noise attenuation is drawn according to the time domain seismic signal after random noise attenuation.
在本实施例中，地震数据获取单元101获取地震数据x(i,t),i＝1,2,…n，其中n为记录道数，本实施例中，获取的地震数据的记录道数n为70，当然本发明不限于此，可以根据实际情况选择其他数值。In this embodiment, the seismic data acquisition unit 101 acquires seismic data x(i, t), i=1, 2,...n, where n is the number of records, and in this embodiment, the number of records of the acquired seismic data n is 70, and of course the present invention is not limited thereto, and other values can be selected according to actual conditions.
在本实施例中，频率-空间域的地震数据生成单元102将所述地震数据进行傅立叶变换，生成频率-空间域的地震数据，包括：In this embodiment, the seismic data generation unit 102 in the frequency-space domain performs Fourier transform on the seismic data to generate seismic data in the frequency-space domain, including:
对地震数据x(i,t),i＝1,2,…n进行傅立叶变换，生成频率-空间域的地震数据X(i,f),i＝1,2,…n；其中，Perform Fourier transform on seismic data x(i,t), i=1,2,...n to generate frequency-space domain seismic data X(i,f),i=1,2,...n; where,
X(i,f)=∫x(i,t)e-j2πftdt。X(i,f)=∫x(i,t)e-j2πft dt.
在本实施例中，模态分量生成单元103在空间方向上对所述频率-空间域的地震数据进行复数经验模态分解，生成多个模态分量，包括：In this embodiment, the modal component generation unit 103 performs complex empirical mode decomposition on the seismic data in the frequency-space domain in the spatial direction to generate multiple modal components, including:
对所述频率-空间域的地震数据X(i,f),i＝1,2,…n进行复数经验模态分解，生成多个模态分量Cj(i,f),j=1,2,…,m，其中，m是经过经验模态分解的模态分量的个数，在本实施例中，经过经验模态分解的模态分量的个数m为9。performing complex empirical mode decomposition on the seismic data X(i,f) in the frequency-space domain, i=1,2,...n, to generate multiple modal components Cj (i,f),j=1, 2, . . . , m, wherein, m is the number of modal components undergoing empirical mode decomposition, and in this embodiment, the number m of modal components undergoing empirical mode decomposition is 9.
在本实施例中，重构算子生成单元104根据所述频率-空间域的地震数据与所述多个模态分量，利用最优化方法生成自适应信号重构算子，包括：In this embodiment, the reconstruction operator generation unit 104 uses an optimization method to generate an adaptive signal reconstruction operator according to the seismic data in the frequency-space domain and the multiple modal components, including:
步骤1，根据所述频率-空间域的地震数据X(i,f),i＝1,2,…n构建向量X，根据所述多个模态分量Cj(i,f),j=1,2,…,m构建矩阵C，根据所述自适应信号重构算子Aj(f),j=1,2,…,m构建向量A；Step 1, construct a vector X according to the seismic data X(i,f) in the frequency-space domain, i=1,2,...n, and according to the multiple modal components Cj (i,f),j= 1, 2,..., m constructing a matrix C, constructing a vector A according to the adaptive signal reconstruction operator Aj (f), j=1, 2,..., m;
步骤2，建立目标函数E=X-C×AT；Step 2, establish the objective function E=XC×AT ;
步骤3，利用最优化方法计算所述目标函数的最小值，生成所述目标函数最小值下的自适应信号重构算子Aj(f),j=1,2,…,m；Step 3, using an optimization method to calculate the minimum value of the objective function, and generate an adaptive signal reconstruction operator Aj (f),j=1,2,...,m under the minimum value of the objective function;
此时，A=(CCT+βI)-1CTX，其中，I为单位矩阵，β为稳定性阻尼系数，且β=0.01。At this time, A=(CCT +βI)-1 CT X, where I is the identity matrix, β is the stability damping coefficient, and β=0.01.
在本实施例中，频率域地震信号生成单元105根据所述自适应信号重构算子和所述多个模态分量，重构生成频率域地震信号，包括：In this embodiment, the frequency-domain seismic signal generation unit 105 reconstructs and generates frequency-domain seismic signals according to the adaptive signal reconstruction operator and the multiple modal components, including:
利用所述自适应信号重构算子Aj(f),j=1,2,…,m，由所述多个模态分量Cj(i,f),j=1,2,…,m重构所述频率域地震信号S(i,f),i＝1,2,…n：Using the adaptive signal reconstruction operator Aj (f),j=1,2,...,m, from the multiple modal components Cj (i,f),j=1,2,..., m reconstructing the frequency domain seismic signal S(i, f), i=1, 2,...n:
在本实施例中，时间域地震信号生成单元106将所述频率域地震信号进行傅立叶反变换，生成随机噪声衰减后的时间域地震信号，包括：In this embodiment, the time-domain seismic signal generation unit 106 performs inverse Fourier transform on the frequency-domain seismic signal to generate a time-domain seismic signal after random noise attenuation, including:
对重构后得到的所述频率域地震信号进行傅立叶反变换，生成随机噪声衰减后的时间域地震信号s(i,t),i＝1,2,…n：For the frequency domain seismic signal obtained after reconstruction Perform inverse Fourier transform to generate time-domain seismic signals s(i,t) after random noise attenuation, i=1,2,...n:
s(i,t)=∫S(i,f)ej2πftdf。s(i,t)=∫S(i,f)ej2πft df.
在本实施例中，地震剖面图像绘制单元107根据常用的地震数据绘制软件就可以绘制噪声衰减后的地震剖面图像。In this embodiment, the seismic section image drawing unit 107 can draw the noise-attenuated seismic section image according to common seismic data drawing software.
实施例1：Example 1:
本实施例为某油田A区块的应用实例，该勘探区块位于沙漠腹地，沙丘散射在地震数据上产生了强烈的随机干扰，勘探目的层为地下7000米左右的碳酸盐岩储层，碳酸盐岩内幕波阻抗差异较小，反射信号较弱。碳酸盐岩内幕反射完全淹没在散射噪声之中。图5是本勘探区块野外采集的地震记录，在该地震记录上几乎见不到有效反射的影子，很难对碳酸盐岩内幕的结构特征进行分析和描述。图6是利用工业界f-x域空间预测滤波方法进行随机噪声衰减之后的结果，随机噪声得到一定程度的压制，在去噪之后的地震剖面上可以见到时断时续的有效信号。图7是利用本发明的衰减方法进行随机噪声衰减之后的结果，随机噪声得到了更加彻底的衰减，较好地恢复了碳酸盐岩内幕结构的反射特征，为碳酸盐岩构造解释和储层预测提供了高质量的基础数据。This embodiment is an application example of block A of an oil field. The exploration block is located in the hinterland of the desert. The scattering of sand dunes produces strong random interference on the seismic data. The exploration target layer is a carbonate rock reservoir about 7000 meters underground. The difference in wave impedance inside the carbonate rock is small, and the reflection signal is weak. Carbonate interior reflections are completely lost in scattering noise. Figure 5 shows the seismic records collected in the field in this exploration block. There are almost no effective reflection shadows on the seismic records, and it is difficult to analyze and describe the structural characteristics of the carbonate rocks. Figure 6 is the result of random noise attenuation using the f-x domain spatial prediction filtering method in the industry. The random noise is suppressed to a certain extent, and intermittent effective signals can be seen on the seismic section after denoising. Fig. 7 is the result of random noise attenuation using the attenuation method of the present invention. The random noise has been more thoroughly attenuated, and the reflection characteristics of the inner structure of carbonate rocks have been better restored, which is useful for carbonate rock structure interpretation and storage. Stratum forecasts provide high-quality underlying data.
实施例2：Example 2:
本实施例为某油田B区块的应用实例，该区块与A区块相邻，但信噪比略高于A区块的地震数据。图8是该区快采集的地震数据，3100ms之上为砂岩地层，3100ms之下为碳酸盐岩地层，由于随机噪声的污染，在图8所示的地震剖面上只能追踪到几个强反射界面产生的同相轴，弱反射信号完全淹没在噪声干扰之中。图9是利用本发明的衰减方法进行随机噪声衰减之后的结果，随机干扰得到了有效压制，很好地恢复了被随机噪声淹没的弱反射信号，清晰地展示了地层结构及其接触关系，大幅度提高了利用地震信号探测地下结构的精度。This embodiment is an application example of block B in an oil field, which is adjacent to block A, but the signal-to-noise ratio is slightly higher than the seismic data of block A. Figure 8 shows the fast-acquired seismic data in this area. Above 3100ms is the sandstone formation, and below 3100ms is the carbonate formation. Due to random noise pollution, only a few strong The event produced by the reflection interface, the weak reflection signal is completely submerged in the noise interference. Fig. 9 is the result of random noise attenuation using the attenuation method of the present invention, the random interference is effectively suppressed, the weak reflection signal submerged by the random noise is well restored, and the stratum structure and its contact relationship are clearly shown. Amplitude improves the accuracy of detecting subsurface structures using seismic signals.
本发明实施例的衰减地震数据随机噪声的方法与系统，可以自动识别并重构被随机噪声严重污染的地震信号，有效压制随机噪声对地震信号的影响，提高了地震资料信噪比，增强了地震信号对地下复杂结构和油气储层的检测精度，为地震资料构造解释和储层预测提供了高质量的基础数据。The method and system for attenuating random noise in seismic data according to the embodiments of the present invention can automatically identify and reconstruct seismic signals seriously polluted by random noise, effectively suppress the influence of random noise on seismic signals, improve the signal-to-noise ratio of seismic data, and enhance the The detection accuracy of seismic signals for complex underground structures and oil and gas reservoirs provides high-quality basic data for structural interpretation of seismic data and reservoir prediction.
以上所述的具体实施例，对本发明的目的、技术方案和有益效果进行了进一步详细说明，所应理解的是，以上所述仅为本发明的具体实施例而已，并不用于限定本发明的保护范围，凡在本发明的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。The specific embodiments described above have further described the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above descriptions are only specific embodiments of the present invention and are not intended to limit the scope of the present invention. Protection scope, within the spirit and principles of the present invention, any modification, equivalent replacement, improvement, etc., shall be included in the protection scope of the present invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210483278.XACN102998706B (en) | 2012-11-23 | 2012-11-23 | Method and system for attenuating seismic data random noise |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210483278.XACN102998706B (en) | 2012-11-23 | 2012-11-23 | Method and system for attenuating seismic data random noise |
| Publication Number | Publication Date |
|---|---|
| CN102998706A CN102998706A (en) | 2013-03-27 |
| CN102998706Btrue CN102998706B (en) | 2015-05-20 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210483278.XAActiveCN102998706B (en) | 2012-11-23 | 2012-11-23 | Method and system for attenuating seismic data random noise |
| Country | Link |
|---|---|
| CN (1) | CN102998706B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103308940B (en)* | 2013-05-27 | 2016-08-10 | 中山大学 | The empirical mode decomposition method of seismic profile |
| CN104375184B (en)* | 2013-08-15 | 2017-04-05 | 中国石油天然气集团公司 | A kind of efficient geological data random noise attenuation method |
| CN105372707A (en)* | 2014-08-27 | 2016-03-02 | 中国石油化工股份有限公司 | Method for attenuating multi-scale seismic data random noise |
| CN106249284B (en)* | 2015-06-15 | 2019-01-08 | 中国石油化工股份有限公司 | Deamplification decomposition method based on the reflection of Q value difference |
| CN105116442B (en)* | 2015-07-24 | 2019-01-01 | 长江大学 | The reconstructing method of the weak seismic reflection signals of lithologic deposit |
| CN106483563B (en)* | 2015-08-25 | 2018-10-16 | 中国石油天然气股份有限公司 | Seismic energy compensation method based on complementary set empirical mode decomposition |
| CN106569274B (en)* | 2015-10-10 | 2019-01-08 | 中国石油化工股份有限公司 | The method and apparatus for removing the random noise in seismic signal |
| CN105510969B (en)* | 2016-01-17 | 2017-12-15 | 中国船舶重工集团公司第七六○研究所 | A kind of method for judging navigation ship seismic-wave field signal line spectrum feature |
| CN106199716A (en)* | 2016-06-21 | 2016-12-07 | 中国石油集团川庆钻探工程有限公司地球物理勘探公司 | A kind of post-stack seismic data stochastic noise suppression method |
| CN106526665B (en)* | 2016-09-20 | 2018-07-13 | 中国石油天然气股份有限公司 | Method and device for random noise attenuation |
| CN107067374B (en)* | 2016-12-23 | 2019-11-01 | 西安电子科技大学 | Two-stage interpolation-based matrix completion two-dimensional image processing method |
| CN107024718A (en)* | 2017-05-31 | 2017-08-08 | 西南石油大学 | Poststack earthquake fluid Forecasting Methodology based on CEEMD SPWVD Time-frequency Spectrum Analysis |
| CN107526103B (en)* | 2017-08-04 | 2019-08-02 | 中国石油集团东方地球物理勘探有限责任公司 | The acquiring method of Processing Seismic Data and its threshold and useful signal frequency |
| CN112799131B (en)* | 2019-11-13 | 2024-05-28 | 中国石油天然气股份有限公司 | Seismic data denoising method and device |
| CN112232321B (en)* | 2020-12-14 | 2021-03-19 | 西南交通大学 | Vibration data interference noise reduction method, device and equipment and readable storage medium |
| CN113031074B (en)* | 2021-03-01 | 2021-12-07 | 中国科学院地质与地球物理研究所 | Logging curve denoising method, device, equipment and storage medium |
| CN115685313B (en)* | 2021-07-21 | 2025-08-29 | 中国石油天然气股份有限公司 | A method for recovering weak reflected signals by eliminating noise interference |
| CN114114415A (en)* | 2021-11-24 | 2022-03-01 | 中铁西南科学研究院有限公司 | Ultra-long distance passive source geological forecasting method for drilling and blasting construction tunnel |
| CN114627167B (en)* | 2022-02-25 | 2025-05-30 | 广州瑞多思医疗科技有限公司 | Arbitrary modality image registration method and device based on neural network |
| CN116068624B (en)* | 2023-02-24 | 2025-05-13 | 西安交通大学 | Seismic random noise suppression method and system based on reduced-rank F-X prediction filtering |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7464006B1 (en)* | 2003-10-09 | 2008-12-09 | The United States Of America As Represented By The Administrator Of The National Aeronautics And Space Administration | Analyzing nonstationary financial time series via hilbert-huang transform (HHT) |
| CN102520396A (en)* | 2012-01-04 | 2012-06-27 | 西安电子科技大学 | Time-varying Narrowband Interference Suppression Method Based on Complex Empirical Mode Decomposition |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7464006B1 (en)* | 2003-10-09 | 2008-12-09 | The United States Of America As Represented By The Administrator Of The National Aeronautics And Space Administration | Analyzing nonstationary financial time series via hilbert-huang transform (HHT) |
| CN102520396A (en)* | 2012-01-04 | 2012-06-27 | 西安电子科技大学 | Time-varying Narrowband Interference Suppression Method Based on Complex Empirical Mode Decomposition |
| Title |
|---|
| 《Bivariate Empirical Mode Decomposition》;Gabriel Rilling et al.;《IEEE SIGNAL PROCESSING LETTERS》;20071231;第14卷(第12期);第1-10页* |
| 《Random and coherent noise attenuation by empirical mode decomposition》;Maïza Bekara et al.;《GEOPHYSICS》;20090414;第74卷(第5期);第90-92页* |
| 《基于经验模态分解的f-x域噪声衰减方法研究》;王峣钧等;《中国地球物理学会第二十七届年会论文集》;20120625;第551页* |
| 周开明,等.《基于经验模态分解的去噪方法研究》.《江汉石油科技》.2011,第21卷(第1期),第11-14页.* |
| Publication number | Publication date |
|---|---|
| CN102998706A (en) | 2013-03-27 |
| Publication | Publication Date | Title |
|---|---|---|
| CN102998706B (en) | Method and system for attenuating seismic data random noise | |
| Gómez et al. | A simple method inspired by empirical mode decomposition for denoising seismic data | |
| Pinnegar et al. | The S-transform with windows of arbitrary and varying shape | |
| CN108983286B (en) | A Seismic Data Denoising Method Combined with CEEMD and Generalized S Transform | |
| CN112946749B (en) | Method for suppressing seismic multiples based on data augmentation training deep neural network | |
| CN104395779B (en) | System and method for estimating and weakening the noise in geological data | |
| CN108267784A (en) | A kind of seismic signal random noise compression process method | |
| CN104536044B (en) | The interpolation denoising method and system of a kind of geological data | |
| CN107144880B (en) | A Seismic Wave Field Separation Method | |
| CN106443787B (en) | Prestack seismic gather presses method for de-noising and its device | |
| CN103399348B (en) | Based on the Denoising of Seismic Data method of Shearlet conversion | |
| CN111158051B (en) | A Joint Constrained Random Noise Suppression Method Based on Sparse Regularization | |
| CN101482617A (en) | Synthetic aperture radar image denoising method based on non-down sampling profile wave | |
| Li et al. | Wavelet-based higher order correlative stacking for seismic data denoising in the curvelet domain | |
| CN107144879A (en) | A kind of seismic wave noise-reduction method combined based on adaptive-filtering with wavelet transformation | |
| CN111337980B (en) | Carbon dioxide storage monitoring method and system based on time-lapse full waveform inversion | |
| Du et al. | Study on optical fiber gas-holdup meter signal denoising using improved threshold wavelet transform | |
| CN108828658A (en) | A kind of ocean bottom seismic data reconstructing method | |
| CN119395764B (en) | Random noise denoising method for deep reflection seismic data based on multi-scale feature fusion | |
| Saad et al. | Signal enhancement in distributed acoustic sensing data using a guided unsupervised deep learning network | |
| Huang et al. | Frequency–space-dependent smoothing regularized nonstationary predictive filtering | |
| Shang et al. | Seismic data analysis using synchrosqueezing wavelet transform | |
| Zheng et al. | The surface wave suppression using the second generation curvelet transform | |
| CN119090723A (en) | A super-resolution reconstruction method for post-stack seismic profiles | |
| Fogat et al. | A Comparative Analysis of Convolutional Neural Networks for Seismic Noise Attenuation |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |