CN115184999B - Marchenko imaging focusing function correction method based on deep learning - Google Patents
Marchenko imaging focusing function correction method based on deep learning Download PDFInfo
- Publication number
- CN115184999B CN115184999B CN202210796555.6A CN202210796555A CN115184999B CN 115184999 B CN115184999 B CN 115184999B CN 202210796555 A CN202210796555 A CN 202210796555A CN 115184999 B CN115184999 B CN 115184999B
- Authority
- CN
- China
- Prior art keywords
- focusing function
- model
- marchenko
- imaging
- function
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000006870 function Effects 0.000 title claims abstract description 93
- 238000003384 imaging method Methods 0.000 title claims abstract description 38
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000012937 correction Methods 0.000 title claims abstract description 17
- 238000013135 deep learning Methods 0.000 title claims abstract description 17
- 238000012549 training Methods 0.000 claims abstract description 15
- 238000013526 transfer learning Methods 0.000 claims abstract description 10
- 230000004044 response Effects 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000009499 grossing Methods 0.000 claims description 3
- 238000004613 tight binding model Methods 0.000 claims 2
- 238000013508 migration Methods 0.000 claims 1
- 230000005012 migration Effects 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 8
- 101001121408 Homo sapiens L-amino-acid oxidase Proteins 0.000 description 2
- 102100026388 L-amino-acid oxidase Human genes 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 101000827703 Homo sapiens Polyphosphoinositide phosphatase Proteins 0.000 description 1
- 102100023591 Polyphosphoinositide phosphatase Human genes 0.000 description 1
- 101100012902 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) FIG2 gene Proteins 0.000 description 1
- 101100233916 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) KAR5 gene Proteins 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000011229 interlayer Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
- G01V1/28—Processing seismic data, e.g. for interpretation or for event detection
- G01V1/36—Effecting static or dynamic corrections on records, e.g. correcting spread; Correlating seismic signals; Eliminating effects of unwanted energy
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Acoustics & Sound (AREA)
- Environmental & Geological Engineering (AREA)
- Geology (AREA)
- General Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Geophysics (AREA)
- Radar Systems Or Details Thereof (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
本发明适用于地震勘探技术领域，提供了基于深度学习的Marchenko成像聚焦函数校正方法，其步骤包括：将观测系统得到的地震数据和背景模型得到的直达波作为输入，经过Marchenko方程得到的聚焦函数作为需要校正的原训练数据；将参考模型得到的精确聚焦函数作为标签数据，按照设计的U‑net架构训练模型；训练好的模型用来校正含误差的聚焦函数，最终得到更精确的Marchenko成像。该方法可有效的在未知区域校正聚焦函数，从而加速Marchenko成像过程，同时通过迁移学习，模型可以有效地校正直达波含走时误差和地震数据有缺失信息的情况下的聚焦函数，而不需要花费大量时间去重建丢失的输入数据，节省了时间和成本。
The present invention is applicable to the field of seismic exploration technology, and provides a Marchenko imaging focusing function correction method based on deep learning, the steps of which include: using the seismic data obtained by the observation system and the direct wave obtained by the background model as input, and the focusing function obtained by the Marchenko equation as the original training data to be corrected; using the precise focusing function obtained by the reference model as label data, and training the model according to the designed U-net architecture; the trained model is used to correct the focusing function containing errors, and finally obtain more accurate Marchenko imaging. This method can effectively correct the focusing function in unknown areas, thereby accelerating the Marchenko imaging process. At the same time, through transfer learning, the model can effectively correct the focusing function when the direct wave contains travel time errors and the seismic data has missing information, without spending a lot of time to rebuild the lost input data, saving time and cost.
Description
技术领域Technical Field
本发明属于地震勘探技术领域，尤其涉及基于深度学习的Marchenko成像聚焦函数校正方法。The present invention belongs to the technical field of seismic exploration, and in particular to a Marchenko imaging focusing function correction method based on deep learning.
背景技术Background technique
地震勘探中往往会遇到上覆盖层地质结构复杂的情况。震源从地表激发，经过复杂的上覆盖层会使地震波扭曲变形，从而对接收器接收到的地震记录产生严重影响。传统的虚震源法可以绕开上覆盖层，将地表震源重建到地下接收器位置。这一方法需要地下有实际的接收器。然而，在实际过程中可能无法将接收器放在地下深处的指定位置。Marchenko成像方法基于“自动聚焦”的概念，仅需要地表反射响应和地下焦点到地表的直达波就能计算出焦点到地表的格林函数，也就是说，相对于虚震源法而言，该方法不需要在地下放置实际接收器，并且可以指定地下任意位置(即，焦点处)作为虚拟接收点。由此得到的格林函数作为成像的输入，最终得到的地下成像抑制了和层间多次波相关的伪影。相较于其它传统成像方法，Marchenko成像方法使用的格林函数能更精确地解释多次波。In seismic exploration, we often encounter situations where the geological structure of the overburden is complex. When the earthquake source is excited from the surface, the complex overburden will distort the seismic waves, which will have a serious impact on the seismic records received by the receiver. The traditional virtual source method can bypass the overburden and reconstruct the surface earthquake source to the underground receiver position. This method requires an actual receiver underground. However, in practice, it may not be possible to place the receiver at a specified position deep underground. The Marchenko imaging method is based on the concept of "autofocus". It only requires the surface reflection response and the direct wave from the underground focus to the surface to calculate the Green's function from the focus to the surface. That is to say, compared with the virtual source method, this method does not require the placement of actual receivers underground, and any underground position (i.e., the focus) can be specified as a virtual receiving point. The Green's function obtained in this way is used as the input of imaging, and the final underground imaging suppresses artifacts related to interlayer multiple waves. Compared with other traditional imaging methods, the Green's function used in the Marchenko imaging method can more accurately interpret multiple waves.
常规的Marchenko方法使用的背景模型是估计得出的，不精确的背景模型影响到直达波的估计，从而使迭代求得的聚焦函数产生误差，最终导致成像聚焦在错误的位置。此外，走时误差和不完全采样得到的地表反射响应，都会导致聚焦函数和后续的成像出现错误。The background model used in the conventional Marchenko method is estimated. An inaccurate background model affects the estimation of the direct wave, which causes errors in the iterative focusing function, and ultimately causes the imaging to focus on the wrong position. In addition, travel time errors and incompletely sampled surface reflection responses can cause errors in the focusing function and subsequent imaging.
发明内容Summary of the invention
本发明实施例的目的在于提供基于深度学习的Marchenko成像聚焦函数校正方法，旨在解决常规的Marchenko方法使用的背景模型是估计得出的，不精确的背景模型影响到直达波的估计，从而使迭代求得的聚焦函数产生误差，最终导致成像聚焦在错误的位置。此外，走时误差和不完全采样得到的地表反射响应，都会导致聚焦函数和后续的成像出现错误的问题。The purpose of the embodiment of the present invention is to provide a Marchenko imaging focus function correction method based on deep learning, aiming to solve the problem that the background model used in the conventional Marchenko method is estimated, and the inaccurate background model affects the estimation of the direct wave, thereby causing errors in the iteratively obtained focus function, and ultimately causing the imaging to focus on the wrong position. In addition, travel time errors and surface reflection responses obtained by incomplete sampling will cause errors in the focus function and subsequent imaging.
本发明实施例是这样实现的，基于深度学习的Marchenko成像聚焦函数校正方法，包括以下步骤：The embodiment of the present invention is implemented as follows: a Marchenko imaging focusing function correction method based on deep learning includes the following steps:
步骤1：将观测系统得到的地震数据和背景模型得到的直达波作为输入，经过Marchenko方程得到的聚焦函数作为需要校正的原训练数据；Step 1: The seismic data obtained by the observation system and the direct wave obtained by the background model are used as input, and the focusing function obtained by the Marchenko equation is used as the original training data that needs to be corrected;
步骤2：将参考模型得到的精确聚焦函数作为标签数据，按照设计的U-net架构训练模型；Step 2: Use the precise focusing function obtained by the reference model as label data and train the model according to the designed U-net architecture;
步骤3：训练好的模型用来校正含误差的聚焦函数，最终得到更精确的Marchenko成像。Step 3: The trained model is used to correct the focusing function containing errors, and finally obtain more accurate Marchenko imaging.
进一步的技术方案，所述步骤1的具体步骤包括：In a further technical solution, the specific steps of step 1 include:
步骤1.1：通过参考模型建立观测系统得到地表反射响应；Step 1.1: Establish an observation system through the reference model to obtain the surface reflection response;
步骤1.2：初始下行聚焦函数用格林函数直接到达的时间反转来近似：Step 1.2: The initial downlink focusing function is approximated by the time reversal of the direct arrival of the Green's function:
即使用平滑模型估计出的指定区域内的焦点到地表的直达波。That is, the direct wave from the focus to the surface in the specified area estimated using a smoothing model.
初始上行聚焦函数为：The initial uplink focusing function is:
步骤1.3：按照迭代方案求解出含误差的聚焦函数：Step 1.3: Solve the focusing function with error according to the iterative scheme:
其中in
k为迭代次数，是第k次迭代的/>的尾波，符号*表示时域卷积。k is the number of iterations, is the kth iteration/> The symbol * represents the time domain convolution.
步骤1.4：通过参考模型计算出精确的直达波，代入步骤1.3，得到精确的聚焦函数。Step 1.4: Calculate the precise direct wave through the reference model and substitute it into step 1.3 to obtain the precise focusing function.
进一步的技术方案，所述步骤2的具体步骤包括：In a further technical solution, the specific steps of step 2 include:
设计U-Net网络架构，将含误差的聚焦函数作为训练数据，精确的聚焦函数作为标签数据进行模型训练。使用Adam优化器，初始学习速率设置为0.0001，批量大小为16。由于本研究要解决的问题是预测具体数值，即回归问题，因此损失函数使用均方根误差(MSE)，其定义如下：The U-Net network architecture is designed, and the focusing function with errors is used as training data, and the accurate focusing function is used as label data for model training. The Adam optimizer is used, the initial learning rate is set to 0.0001, and the batch size is 16. Since the problem to be solved in this study is to predict specific values, that is, regression problems, the loss function uses the root mean square error (MSE), which is defined as follows:
其中，yi是含误差的聚焦函数y中第i元素的标签值，yi'是神经网络的第i元素的预测值，n是该聚焦函数中元素的总数。图4给出了聚焦函数校正前后的对比效果。Among them, yi is the label value of the i-th element in the focusing function y with error, yi ' is the predicted value of the i-th element of the neural network, and n is the total number of elements in the focusing function. Figure 4 shows the comparison effect before and after the focusing function correction.
进一步的技术方案，所述步骤3包括以下具体步骤：Further technical solution, the step 3 comprises the following specific steps:
步骤3.1：将训练好的模型作为预训练模型，对相应区域的聚焦函数进行迁移学习；Step 3.1: Use the trained model as a pre-trained model to perform transfer learning on the focus function of the corresponding area;
步骤3.2：通过Marchenko方程来获得格林函数：Step 3.2: Obtain Green's function through Marchenko equation:
步骤3.3：进行Marchenko成像，成像公式如下：Step 3.3: Perform Marchenko imaging. The imaging formula is as follows:
步骤3.4：将含有走时误差的聚焦函数和不完全采样得到的聚焦函数分别放入模型进行迁移学习，得到校正后的聚焦函数。Step 3.4: Put the focusing function containing travel time error and the focusing function obtained by incomplete sampling into the model for transfer learning to obtain the corrected focusing function.
本发明实施例提供的基于深度学习的Marchenko成像聚焦函数校正方法，引入了深度学习来处理由于直达波的计算错误而导致的聚焦函数估计错误，并且利用了迁移学习使得在更多未知区域校正聚焦函数变得非常有效，从而加速Marchenko成像过程。此外，通过将此方法的应用扩展到Marchenko方法的其他场景：直达波的走时误差导致的时间偏移，以及不完全采样导致的伪影和间隙。通过迁移学习，我们的模型可以有效地校正直达波含走时误差和地震数据有缺失信息的情况下的聚焦函数，而不需要花费大量时间去重建丢失的输入数据，证明了Marchenko方法中深度学习和迁移学习的好处和广泛应用，节省了时间和成本。The deep learning-based Marchenko imaging focusing function correction method provided in an embodiment of the present invention introduces deep learning to deal with focusing function estimation errors caused by calculation errors of direct waves, and utilizes transfer learning to make it very effective to correct the focusing function in more unknown areas, thereby accelerating the Marchenko imaging process. In addition, by extending the application of this method to other scenarios of the Marchenko method: time offset caused by travel time errors of direct waves, and artifacts and gaps caused by incomplete sampling. Through transfer learning, our model can effectively correct the focusing function when the direct wave contains travel time errors and the seismic data has missing information, without spending a lot of time to reconstruct the lost input data, which proves the benefits and wide application of deep learning and transfer learning in the Marchenko method, saving time and cost.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
图1为本发明实施例提供的基于深度学习的Marchenko成像聚焦函数校正方法的算法实现流程图；FIG1 is a flowchart of an algorithm implementation of a Marchenko imaging focusing function correction method based on deep learning provided by an embodiment of the present invention;
图2为本发明实施例提供的基于深度学习的Marchenko成像聚焦函数校正方法中的参考模型图；FIG2 is a diagram of a reference model in a Marchenko imaging focusing function correction method based on deep learning provided by an embodiment of the present invention;
图3为本发明实施例提供的基于深度学习的Marchenko成像聚焦函数校正方法中的网络架构图；FIG3 is a network architecture diagram of a Marchenko imaging focusing function correction method based on deep learning provided by an embodiment of the present invention;
图4为本发明实施例提供的基于深度学习的Marchenko成像聚焦函数校正方法中的聚焦函数对比的结构示意图；FIG4 is a schematic structural diagram of focusing function comparison in a Marchenko imaging focusing function correction method based on deep learning provided by an embodiment of the present invention;
图5为本发明实施例提供的基于深度学习的Marchenko成像聚焦函数校正方法中的成像结果对比图；FIG5 is a comparison diagram of imaging results in a Marchenko imaging focus function correction method based on deep learning provided by an embodiment of the present invention;
图6为本发明实施例提供的基于深度学习的Marchenko成像聚焦函数校正方法中的含走时误差的下行聚焦函数对比图；FIG6 is a comparison diagram of downlink focusing functions containing travel time errors in the Marchenko imaging focusing function correction method based on deep learning provided by an embodiment of the present invention;
图7为本发明实施例提供的基于深度学习的Marchenko成像聚焦函数校正方法中的不完全采样得到的下行聚焦函数对比图。FIG7 is a comparison diagram of the downlink focusing function obtained by incomplete sampling in the Marchenko imaging focusing function correction method based on deep learning provided in an embodiment of the present invention.
具体实施方式Detailed ways
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
以下结合具体实施例对本发明的具体实现进行详细描述。The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.
如图1所示，为本发明一个实施例提供的基于深度学习的Marchenko成像聚焦函数校正方法，包括以下步骤：As shown in FIG1 , a Marchenko imaging focusing function correction method based on deep learning provided in one embodiment of the present invention includes the following steps:
步骤1：将观测系统得到的地震数据和背景模型得到的直达波作为输入，经过Marchenko方程得到的聚焦函数作为需要校正的原训练数据；Step 1: The seismic data obtained by the observation system and the direct wave obtained by the background model are used as input, and the focusing function obtained by the Marchenko equation is used as the original training data that needs to be corrected;
步骤2：将参考模型得到的精确聚焦函数作为标签数据，按照设计的U-net架构训练模型；Step 2: Use the precise focusing function obtained by the reference model as label data and train the model according to the designed U-net architecture;
步骤3：训练好的模型用来校正含误差的聚焦函数，最终得到更精确的Marchenko成像。Step 3: The trained model is used to correct the focusing function containing errors, and finally obtain more accurate Marchenko imaging.
作为本发明的优选实施例，所述步骤1的具体步骤包括：As a preferred embodiment of the present invention, the specific steps of step 1 include:
步骤1.1：通过参考模型建立观测系统得到地表反射响应；Step 1.1: Establish an observation system through the reference model to obtain the surface reflection response;
步骤1.2：初始下行聚焦函数用格林函数直接到达的时间反转来近似：Step 1.2: The initial downlink focusing function is approximated by the time reversal of the direct arrival of the Green's function:
即使用平滑模型估计出的图2中区域A内的焦点到地表的直达波。That is, the direct wave from the focus to the surface in area A in Figure 2 estimated using the smoothing model.
初始上行聚焦函数为：The initial uplink focusing function is:
步骤1.3：按照迭代方案求解出含误差的聚焦函数：Step 1.3: Solve the focusing function with error according to the iterative scheme:
其中in
k为迭代次数，是第k次迭代的/>的尾波，符号*表示时域卷积。k is the number of iterations, is the kth iteration/> The symbol * represents the time domain convolution.
步骤1.4：通过参考模型计算出精确的直达波，代入步骤1.3，得到精确的聚焦函数。Step 1.4: Calculate the precise direct wave through the reference model and substitute it into step 1.3 to obtain the precise focusing function.
作为本发明的优选实施例，所述步骤2的具体步骤包括：As a preferred embodiment of the present invention, the specific steps of step 2 include:
设计U-Net网络架构(图3)，将含误差的聚焦函数作为训练数据，精确的聚焦函数作为标签数据进行模型训练。使用Adam优化器，初始学习速率设置为0.0001，批量大小为16。由于本研究要解决的问题是预测具体数值，即回归问题，因此损失函数使用均方根误差(MSE)，其定义如下：Design the U-Net network architecture (Figure 3), use the focusing function with errors as training data, and the accurate focusing function as label data for model training. Use the Adam optimizer, set the initial learning rate to 0.0001, and the batch size to 16. Since the problem to be solved in this study is to predict specific values, that is, regression problems, the loss function uses the root mean square error (MSE), which is defined as follows:
其中，yi是含误差的聚焦函数y中第i元素的标签值，yi'是神经网络的第i元素的预测值，n是该聚焦函数中元素的总数。图4给出了聚焦函数校正前后的对比效果。Among them, yi is the label value of the i-th element in the focusing function y with error, yi ' is the predicted value of the i- th element of the neural network, and n is the total number of elements in the focusing function. Figure 4 shows the comparison effect before and after the focusing function correction.
作为本发明的优选实施例，所述步骤3包括以下具体步骤：As a preferred embodiment of the present invention, step 3 includes the following specific steps:
步骤3.1：将训练好的模型作为预训练模型，对图2中区域B-F的聚焦函数进行迁移学习；Step 3.1: Use the trained model as a pre-trained model to perform transfer learning on the focus function of area B-F in Figure 2;
步骤3.2：通过Marchenko方程来获得格林函数：Step 3.2: Obtain Green's function through Marchenko equation:
步骤3.3：进行Marchenko成像，成像公式如下：Step 3.3: Perform Marchenko imaging. The imaging formula is as follows:
图5给出了最终成像结果的对比效果。Figure 5 shows the comparison of the final imaging results.
步骤3.4：将含有走时误差的聚焦函数和不完全采样得到的聚焦函数分别放入模型进行迁移学习，得到校正后的聚焦函数。图6和图7展示了迁移学习的训练成果。其中，图6中，a为未经校正的下行聚焦函数，b为校正后的下行聚焦函数，c为参考下行聚焦函数，d为单道放大对比图。图7中，a为未经校正的下行聚焦函数，b为校正后的下行聚焦函数，c为参考下行聚焦函数，d为单道放大对比图。Step 3.4: Put the focusing function containing the travel time error and the focusing function obtained by incomplete sampling into the model for transfer learning respectively to obtain the corrected focusing function. Figures 6 and 7 show the training results of transfer learning. Among them, in Figure 6, a is the uncorrected downlink focusing function, b is the corrected downlink focusing function, c is the reference downlink focusing function, and d is the single-channel magnification comparison diagram. In Figure 7, a is the uncorrected downlink focusing function, b is the corrected downlink focusing function, c is the reference downlink focusing function, and d is the single-channel magnification comparison diagram.
以上所述仅为本发明的较佳实施例而已，并不用以限制本发明，凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明的保护范围之内。The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims (1)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210796555.6A CN115184999B (en) | 2022-07-06 | 2022-07-06 | Marchenko imaging focusing function correction method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210796555.6A CN115184999B (en) | 2022-07-06 | 2022-07-06 | Marchenko imaging focusing function correction method based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115184999A CN115184999A (en) | 2022-10-14 |
| CN115184999B true CN115184999B (en) | 2024-04-19 |
Family
ID=83517394
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210796555.6A Active CN115184999B (en) | 2022-07-06 | 2022-07-06 | Marchenko imaging focusing function correction method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115184999B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118033746B (en) * | 2024-04-07 | 2024-07-09 | 吉林大学 | A target-oriented Marchenko imaging method for ocean streamer seismic data |
| CN118131322B (en) * | 2024-05-06 | 2024-08-13 | 吉林大学 | A method for microseismic source location based on Marchenko imaging |
| CN119416660B (en) * | 2025-01-06 | 2025-05-06 | 吉林大学 | A new 3D earthquake travel-time simulation method based on PIUFNO |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2015560A2 (en) * | 2007-07-13 | 2009-01-14 | Morpho Inc. | Image data processing method and imaging apparatus |
| CN107329233A (en) * | 2017-08-25 | 2017-11-07 | 西南大学 | A kind of droplet type PCR instrument Atomatic focusing method based on neutral net |
| CN113219535A (en) * | 2021-05-21 | 2021-08-06 | 东北石油大学 | One-step method interlayer multiple pressing method based on Marchenko theory |
| CN114428342A (en) * | 2020-10-10 | 2022-05-03 | 中国石油化工股份有限公司 | Markenko Green function reconstruction method and system based on MPI |
-
2022
- 2022-07-06 CN CN202210796555.6A patent/CN115184999B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2015560A2 (en) * | 2007-07-13 | 2009-01-14 | Morpho Inc. | Image data processing method and imaging apparatus |
| CN107329233A (en) * | 2017-08-25 | 2017-11-07 | 西南大学 | A kind of droplet type PCR instrument Atomatic focusing method based on neutral net |
| CN114428342A (en) * | 2020-10-10 | 2022-05-03 | 中国石油化工股份有限公司 | Markenko Green function reconstruction method and system based on MPI |
| CN113219535A (en) * | 2021-05-21 | 2021-08-06 | 东北石油大学 | One-step method interlayer multiple pressing method based on Marchenko theory |
Non-Patent Citations (1)
| Title |
|---|
| 机载合成孔径雷达ETF成像算法研究;王国华;孙进平;;遥测遥控;20060620(第03期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115184999A (en) | 2022-10-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115184999B (en) | Marchenko imaging focusing function correction method based on deep learning | |
| CN110766044B (en) | Neural network training method based on Gaussian process prior guidance | |
| CN102469103B (en) | Trojan event prediction method based on BP (Back Propagation) neural network | |
| CN104794501A (en) | Mode identification method and device | |
| WO2021196528A1 (en) | Global ionospheric total electron content prediction method based on deep recurrent neural network | |
| EP4459504A1 (en) | Data processing method, network training method, electronic device, and storage medium | |
| CN113011475B (en) | Distributed fusion method considering correlated noise and random parameter matrix | |
| CN110083981A (en) | A kind of iterative learning Fault Estimation method under any initial state based on forgetting factor | |
| CN105760896A (en) | Corrosion source joint de-noising method for multi-source heterogeneous big data | |
| KR20230110143A (en) | Label with noise processing method and apparatus of thereof | |
| CN118534510A (en) | Beidou satellite-based intelligent positioning method and system | |
| CN119672158B (en) | Remote sensing optical time sequence image reconstruction method, device, terminal and storage medium | |
| CN118520777B (en) | Mask optimization method, system, device and storage medium based on double diffusion model | |
| CN106959121A (en) | A kind of application process of the learning algorithm in magnetic compass error compensation that transfinite of autonomous reversely tuning | |
| CN113875228B (en) | Video frame inserting method and device and computer readable storage medium | |
| Boudjenouia et al. | Robust, blind multichannel image identification and restoration using stack decoder | |
| CN116859391A (en) | InSAR tropospheric delay correction method based on high-frequency texture cross-correlation | |
| CN110398773B (en) | A Restoration and Reconstruction Method for Partially Missing Seismic Data | |
| CN105182299A (en) | Non-orthogonal joint diagonalization echo signal processing methods based on complex rotation matrix | |
| CN113708913A (en) | Method for judging validity of symmetric halved search estimation value of timing deviation | |
| CN115079257A (en) | Q-value estimation and seismic attenuation compensation method based on fusion network | |
| CN114337745A (en) | A Model-Driven Deep Learning Method for Massive MIMO Signal Detection | |
| Mehrizi et al. | Detection of change points in piecewise polynomial signals using trend filtering | |
| CN108737298A (en) | A kind of SCMA blind checking methods based on image procossing | |
| Miettinen et al. | Robust least mean squares estimation of graph signals |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |