CN113820749A - Seismic data velocity field anomaly inversion method based on machine learning - Google Patents
Seismic data velocity field anomaly inversion method based on machine learning Download PDFInfo
- Publication number
- CN113820749A CN113820749A CN202111118628.8A CN202111118628A CN113820749A CN 113820749 A CN113820749 A CN 113820749A CN 202111118628 A CN202111118628 A CN 202111118628A CN 113820749 A CN113820749 A CN 113820749A
- Authority
- CN
- China
- Prior art keywords
- velocity field
- image
- seismic
- net
- seismic data
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000010801 machine learning Methods 0.000 title claims abstract description 15
- 230000002159 abnormal effect Effects 0.000 claims abstract description 30
- 238000013135 deep learning Methods 0.000 claims abstract description 21
- 238000012549 training Methods 0.000 claims abstract description 15
- 238000005070 sampling Methods 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 6
- 238000012805 post-processing Methods 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims 1
- 238000003384 imaging method Methods 0.000 abstract description 7
- 230000005012 migration Effects 0.000 abstract description 4
- 238000013508 migration Methods 0.000 abstract description 4
- 238000011160 research Methods 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 10
- 238000013527 convolutional neural network Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 230000001419 dependent effect Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
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
- G01V1/30—Analysis
- G01V1/303—Analysis for determining velocity profiles or travel times
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V2210/00—Details of seismic processing or analysis
- G01V2210/60—Analysis
- G01V2210/62—Physical property of subsurface
- G01V2210/622—Velocity, density or impedance
- G01V2210/6222—Velocity; travel time
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Remote Sensing (AREA)
- Acoustics & Sound (AREA)
- Environmental & Geological Engineering (AREA)
- Geology (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geophysics (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a seismic data velocity field anomaly inversion method based on machine learning, which comprises the following steps of: building a U-Net deep learning network facing seismic data and training the U-Net deep learning network by using a training data set, wherein the training data set consists of three types of seismic oscillograms, velocity field images and abnormal point images; and inputting the generated seismic oscillogram into a U-Net deep learning network, generating a velocity field image and an abnormal point image, and predicting a feature tag set. According to the seismic data velocity field anomaly inversion method based on machine learning, a U-Net deep learning network is established, an obtained seismic wave graph is input into the U-Net deep learning network, a velocity field image and an anomaly point image predicted by the method are generated, compared with an actual velocity field and an actual anomaly point, the method has high precision, and research is carried out on shot gather seismic data before migration imaging, so that errors caused by a migration imaging algorithm can be effectively avoided.
Description
Technical Field
The invention relates to the technical field of geophysical, in particular to a seismic data velocity field anomaly inversion method based on machine learning.
Background
Inversion refers to deducing underground geologic structure characteristics, occurrence states of geologic bodies and physical parameters according to the change characteristics of a geophysical field, and chromatographic inversion and full waveform inversion are two most commonly used methods for modeling a velocity field at present.
In practical application, the tomographic inversion method has higher precision for background velocity field modeling, but has a general inversion effect for velocity field anomalies corresponding to a complex geological structure, and high-precision velocity field anomalies are often required to be artificially introduced in the velocity field updating process to accelerate iterative convergence; the full waveform inversion method is proposed in the eighties of the last century, and the method is widely applied in recent years along with the great improvement of cluster computing capacity, so that the full waveform inversion method becomes an effective high-resolution speed field modeling method. Due to the multi-solution nature of the velocity field model, as an iterative algorithm based on data fitting, the convergence of the full waveform inversion process is highly dependent on the accuracy of the velocity field initial model. When the initial model, especially the abnormal value error therein is large, a large phase difference is generated between the synthetic seismic record and the actual seismic record, and when the phase difference is larger than half of the wavelet period of the seismic source, a period jump phenomenon is generated, so that the algorithm converges to a local minimum point to generate an error inversion result. The traditional speed field modeling means generally processes data by using an offset imaging algorithm, which requires the offset imaging algorithm to have higher precision, but in actual application, the precision of abnormal distribution of the speed field is influenced due to errors of the offset imaging algorithm.
Meanwhile, in practical application, the speed field abnormity pickup method usually needs a large amount of manual intervention, the speed field abnormity is introduced by utilizing the past accumulated experience and through a manual modification mode of processing and interpreters, the mode is excessively dependent on the experience and technical level of technicians, the modeling quality is unstable, and the modeling efficiency is low.
Disclosure of Invention
The invention aims to solve the defects in the prior art, and provides a seismic data velocity field anomaly inversion method based on machine learning.
In order to solve the technical problems, the invention provides the following technical scheme: a seismic data velocity field anomaly inversion method based on machine learning comprises the following steps:
(1) building a U-Net deep learning network facing seismic data and training the U-Net deep learning network by using a training data set, wherein the training data set consists of three types of seismic oscillograms, velocity field images and abnormal point images;
(2) inputting the generated seismic oscillogram into a U-Net deep learning network, generating a velocity field image and an abnormal point image, and predicting a feature tag set;
(3) and (3) performing post-processing on the generated image based on a CNN algorithm (convolutional neural network) to further reduce errors, wherein the image needing the post-processing comprises a velocity field image and an abnormal point image.
Preferably, the U-Net deep learning network structure mainly comprises a convolutional layer, a down-sampling, an up-sampling and a ReLU nonlinear activation function, and the U-Net network also has a characteristic channel while adopting an up-sampling and down-sampling method.
Preferably, the U-Net network generates the velocity field image and the abnormal point image simultaneously, and the error function is a weighted sum of errors of the generated velocity field and the abnormal point.
Preferably, the process that the U-Net network is directly mapped to the right end of the U-Net by the image before the down sampling for the first time is cancelled.
The invention has the following beneficial effects:
1. according to the seismic data velocity field anomaly inversion method based on machine learning, a U-Net deep learning network is established, an obtained seismic wave graph is input into the U-Net deep learning network, a velocity field image and an anomaly point image predicted by the method are generated, compared with an actual velocity field and an actual anomaly point, the method has high precision, and research is carried out on shot gather seismic data before migration imaging, so that errors caused by a migration imaging algorithm can be effectively avoided.
2. According to the seismic data velocity field anomaly inversion method based on machine learning, higher-precision velocity field anomaly distribution can be provided at the initial stage of velocity field modeling, convergence of a modeling iteration process can be accelerated, modeling efficiency is improved, a velocity field anomaly model is established in an automatic identification mode by means of a deep learning algorithm, dependence on people is greatly reduced, and errors caused by manual intervention are reduced.
3. The seismic data velocity field anomaly inversion method based on machine learning has the characteristics of data self-learning and iteration, continuous optimization of a deep learning network in a technical process is realized by continuously expanding a training data set, and thus training data can be continuously accumulated to continuously exert value.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
FIG. 1 is a schematic flow diagram of an embodiment of the present invention;
FIG. 2 is a schematic diagram of a comparison between a predicted image and a real image at a single outlier in a normal velocity field according to an embodiment of the present invention;
FIG. 3 is a diagram illustrating a comparison between a predicted image and a real image under a normal velocity field and two outliers according to an embodiment of the present invention;
fig. 4 is a schematic diagram of the comparison between a predicted image and a real image under a single outlier with a fault velocity field according to the embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Examples
As shown in fig. 1-4, the invention provides a seismic data velocity field anomaly inversion method based on machine learning, which comprises the following steps:
(1) building a U-Net deep learning network facing seismic data and training the U-Net deep learning network by using a training data set, wherein the training data set consists of three types of seismic oscillograms, velocity field images and abnormal point images;
the U-Net deep learning network structure mainly comprises a convolution layer, a down sampling layer, an up sampling layer and a ReLU nonlinear activation function, and the U-Net network adopts the up sampling method and the down sampling method and simultaneously has a characteristic channel, so that more information of an original image or down sampling texture can be transmitted in a high-resolution layer.
The U-Net network simultaneously generates a speed field image and an abnormal point image, and the error function of the U-Net network is the weighted sum of the generated speed field and the abnormal point error.
The process that the U-Net network is directly mapped to the right end of the U-Net from the image before down-sampling for the first time is cancelled, and the situation that the generated velocity field contains the outline of the seismic oscillogram can be avoided
(2) Inputting the generated seismic oscillogram into a U-Net deep learning network, generating a velocity field image and an abnormal point image, and predicting a feature tag set;
(3) and (3) performing post-processing on the generated image based on a CNN algorithm (convolutional neural network) to further reduce errors, wherein the image needing the post-processing comprises a velocity field image and an abnormal point image.
After a speed field and an abnormal point image generated by U-Net training are obtained, a convolution kernel with the size of 2 multiplied by 2 is used for gradually traversing the whole image to extract features, and a smoothed image is generated after pooling and full-connection operations.
In summary, when the method is actually used for predicting the seismic velocity field and the abnormal points, only the detector is required to receive data and then process the data, the seismic oscillogram is input to the U-Net, the seismic velocity field and the abnormal points are predicted by the method, in the experiment for predicting the seismic velocity field and the abnormal points, the experiment is divided into a plurality of groups for verification, the data are processed after the detector receives the data after the detector is shot at the shot point, fig. 2-4 show that the prediction results when the normal velocity field, the fault velocity field, the single shot point and the multiple shot points are received and compared with the actual velocity field and the abnormal points, the seismic waveforms are shot at the surface and the surface detectors, and then the underground velocity distribution model is obtained through inversion, and the multiple shot points can improve the covering times of related bins and improve the resolution.
Fig. 2 shows the prediction in the case of a normal velocity field and a single anomaly point, as shown in the figure, the velocity field is divided into five layers, only one anomaly point is provided, the velocity field and the anomaly point are respectively predicted in the case of a single shot point and a plurality of shot points, the effect of predicting the velocity field and the anomaly point is shown in the figure, and the prediction precision of the single shot point and the plurality of shot points is high in the case of the normal velocity field and the single anomaly point; fig. 3 shows the prediction under the normal velocity field and two abnormal points, as shown in the figure, the velocity field is divided into two layers, there are two abnormal points, the prediction is performed when a single shot point and a plurality of shot points are respectively performed, the effect of the predicted velocity field and the abnormal points is shown in the figure, and the result shows that the prediction precision of the single shot point and the plurality of shot points is still high under the normal velocity field and the plurality of abnormal points; fig. 4 shows that in the case of a fault velocity field, a single abnormal point velocity field is divided into three layers, a fault is formed, an abnormal point is formed, the effects of the predicted velocity field and the abnormal point are shown in the figure, and the result shows that under the fault velocity field and the single abnormal point, the prediction accuracy of a single shot point and multiple running points is still high.
It should be noted that, in this document, terms such as "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (4)
1. A seismic data velocity field anomaly inversion method based on machine learning is characterized by comprising the following steps:
(1) building a U-Net deep learning network facing seismic data and training the U-Net deep learning network by using a training data set, wherein the training data set consists of three types of seismic oscillograms, velocity field images and abnormal point images;
(2) inputting the generated seismic oscillogram into a U-Net deep learning network, generating a velocity field image and an abnormal point image, and predicting a feature tag set;
(3) and post-processing the generated image based on the CNN algorithm to further reduce the error, wherein the image needing post-processing comprises a speed field image and an abnormal point image.
2. The seismic data velocity field anomaly inversion method based on machine learning according to claim 1, characterized in that the U-Net deep learning network structure mainly comprises convolution layers, down sampling, up sampling and ReLU nonlinear activation functions, and the U-Net network has characteristic channels while adopting the up sampling and down sampling methods.
3. The seismic data velocity field anomaly inversion method based on machine learning of claim 1, wherein the U-Net network simultaneously generates a velocity field image and an anomaly point image, and the error function is a weighted sum of errors of the generated velocity field and the anomaly point.
4. The machine learning-based seismic data velocity field anomaly inversion method according to claim 1, wherein the U-Net network is cancelled by directly mapping an image before downsampling to the right end of the U-Net network for the first time.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111118628.8A CN113820749A (en) | 2021-09-24 | 2021-09-24 | Seismic data velocity field anomaly inversion method based on machine learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111118628.8A CN113820749A (en) | 2021-09-24 | 2021-09-24 | Seismic data velocity field anomaly inversion method based on machine learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113820749A true CN113820749A (en) | 2021-12-21 |
Family
ID=78921029
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111118628.8A Pending CN113820749A (en) | 2021-09-24 | 2021-09-24 | Seismic data velocity field anomaly inversion method based on machine learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113820749A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114662427A (en) * | 2022-03-08 | 2022-06-24 | 芯华章科技股份有限公司 | Debugging method and device for logic system design |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110308484A (en) * | 2019-06-11 | 2019-10-08 | 中国石油大学(北京) | A kind of chromatography conversion method and system based on deep learning intelligent screening first arrival |
CN111239802A (en) * | 2020-01-19 | 2020-06-05 | 中国海洋大学 | Deep learning speed modeling method based on seismic reflection waveform and velocity spectrum |
US20200183031A1 (en) * | 2018-12-11 | 2020-06-11 | Exxonmobil Upstream Research Company | Automated seismic interpretation-guided inversion |
CN111562611A (en) * | 2020-04-08 | 2020-08-21 | 山东大学 | Semi-supervised depth learning seismic data inversion method based on wave equation drive |
CN111626355A (en) * | 2020-05-27 | 2020-09-04 | 中油奥博(成都)科技有限公司 | Unet + + convolutional neural network-based seismic data first arrival pickup method |
CN111948711A (en) * | 2020-08-06 | 2020-11-17 | 中海石油(中国)有限公司 | Method and system for extracting low-frequency part of seismic data by using deep learning method |
WO2020231918A1 (en) * | 2019-05-13 | 2020-11-19 | Schlumberger Technology Corporation | Training a machine learning system using hard and soft constraints |
CN112925012A (en) * | 2021-01-26 | 2021-06-08 | 中国矿业大学(北京) | Seismic full-waveform inversion method and device |
-
2021
- 2021-09-24 CN CN202111118628.8A patent/CN113820749A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20200183031A1 (en) * | 2018-12-11 | 2020-06-11 | Exxonmobil Upstream Research Company | Automated seismic interpretation-guided inversion |
WO2020231918A1 (en) * | 2019-05-13 | 2020-11-19 | Schlumberger Technology Corporation | Training a machine learning system using hard and soft constraints |
CN110308484A (en) * | 2019-06-11 | 2019-10-08 | 中国石油大学(北京) | A kind of chromatography conversion method and system based on deep learning intelligent screening first arrival |
CN111239802A (en) * | 2020-01-19 | 2020-06-05 | 中国海洋大学 | Deep learning speed modeling method based on seismic reflection waveform and velocity spectrum |
CN111562611A (en) * | 2020-04-08 | 2020-08-21 | 山东大学 | Semi-supervised depth learning seismic data inversion method based on wave equation drive |
CN111626355A (en) * | 2020-05-27 | 2020-09-04 | 中油奥博(成都)科技有限公司 | Unet + + convolutional neural network-based seismic data first arrival pickup method |
CN111948711A (en) * | 2020-08-06 | 2020-11-17 | 中海石油(中国)有限公司 | Method and system for extracting low-frequency part of seismic data by using deep learning method |
CN112925012A (en) * | 2021-01-26 | 2021-06-08 | 中国矿业大学(北京) | Seismic full-waveform inversion method and device |
Non-Patent Citations (1)
Title |
---|
朱庆棠: "《周围神经缺损修复材料的生物制造与临床评估》", 30 September 2018 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114662427A (en) * | 2022-03-08 | 2022-06-24 | 芯华章科技股份有限公司 | Debugging method and device for logic system design |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
AU2019213369B2 (en) | Non-local memory network for semi-supervised video object segmentation | |
CN110674604B (en) | Transformer DGA data prediction method based on multi-dimensional time sequence frame convolution LSTM | |
JP6969637B2 (en) | Causality analysis methods and electronic devices | |
CN110361778B (en) | Seismic data reconstruction method based on generation countermeasure network | |
CN112541572B (en) | Residual oil distribution prediction method based on convolutional encoder-decoder network | |
CN110471113B (en) | Inversion dynamic correction method and device based on unsteady seismic data and storage medium | |
CN111292308A (en) | Convolutional neural network-based infrared defect detection method for photovoltaic solar panel | |
US20210097371A1 (en) | Generative adversarial network with dynamic capacity expansion for continual learning | |
CN113820749A (en) | Seismic data velocity field anomaly inversion method based on machine learning | |
CN116958131B (en) | Image processing method, device, equipment and storage medium | |
GB2579262A (en) | Space-time memory network for locating target object in video content | |
CN112150497A (en) | Local activation method and system based on binary neural network | |
WO2021135281A1 (en) | Multi-layer feature fusion-based endpoint detection method, apparatus, device, and medium | |
CN117236485B (en) | Method for reducing prediction power supply quantity error of power grid based on CNN-BILSTM algorithm | |
EP4237997A1 (en) | Segmentation models having improved strong mask generalization | |
CN110096790B (en) | Irregular defect magnetic flux leakage signal inversion method based on reinforcement learning | |
CN117171713A (en) | Cross self-adaptive deep migration learning method and system based on bearing service life | |
CN113642232B (en) | Intelligent inversion exploration method for surface waves, storage medium and terminal equipment | |
CN116011338A (en) | Full waveform inversion method based on self-encoder and deep neural network | |
CN116152263A (en) | CM-MLP network-based medical image segmentation method | |
CN109856673B (en) | High-resolution Radon transformation data separation technology based on dominant frequency iterative weighting | |
CN113392845A (en) | Deep learning remote sensing image semantic segmentation method and system based on U-NET | |
CN113989642A (en) | InSAR phase unwrapping method based on fusion multi-scale residual error network | |
CN113223038A (en) | Discrete cosine transform-based mask representation instance segmentation method | |
CN117252788B (en) | Image restoration method, device and storage medium |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20211221 |