CN110441820B - Intelligent interpretation method of geological structure - Google Patents
Intelligent interpretation method of geological structure Download PDFInfo
- Publication number
- CN110441820B CN110441820B CN201910777505.1A CN201910777505A CN110441820B CN 110441820 B CN110441820 B CN 110441820B CN 201910777505 A CN201910777505 A CN 201910777505A CN 110441820 B CN110441820 B CN 110441820B
- Authority
- CN
- China
- Prior art keywords
- neural network
- convolutional neural
- interpretation
- layer
- network model
- 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
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. analysis, for interpretation, for correction
- G01V1/30—Analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- 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
- G06N3/08—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
- G01V2210/64—Geostructures, e.g. in 3D data cubes
Abstract
The invention provides an intelligent interpretation method of geological structure, which relates to the technical field of seismic data analysis and comprises the following steps: and generating a label data volume by using a pre-explanation result of the original seismic amplitude data and by using a method of assigning values to different types of structures, and constructing a convolutional neural network model to predict the geological structure as a training set of the convolutional neural network. The mutual relation between the horizon and different types of structures is considered, so that the horizon interpretation at the structure is clearer, and the geological interpretation of different structures is more accurate.
Description
Technical Field
The invention relates to the technical field of seismic data analysis, in particular to an intelligent interpretation method of geological structure.
Background
With the continuous development of coal resource exploration, the accuracy requirement of seismic data interpretation is higher and higher. Initially, horizon interpretation and structure interpretation are both done by manual interpretation, which is interpreted by the interpreter according to his own geological knowledge and experience, and is not only time-consuming but also lacks certain objectivity.
In order to solve the above problems, a current solution is to identify a fault plane by using a full convolution neural network, acquire seismic amplitude fault slice data from a three-dimensional seismic amplitude data volume, and calibrate a fault position on the slice data by using an artificial interpretation method, wherein the slice data is used as tag data. The method uses slice data as label data, which is not representative, and the interpretation result of the geological structure is not accurate.
Disclosure of Invention
In view of the above, the present invention provides an intelligent interpretation method for geological structure to solve the problems in the prior art.
In a first aspect, there is provided a method for intelligent interpretation of geological formations, the method comprising the steps of:
determining an interpretation result of the construction interpretation according to original seismic amplitude data which are pre-interpreted based on the construction in advance, and generating a three-dimensional data body of the construction interpretation result;
determining a type of the constructed interpretation result based on the constructed interpretation result;
based on the type of the construction interpretation result, assigning a value to the three-dimensional data body of the construction interpretation result to generate a label data body of the construction interpretation result;
establishing a training set according to the label data volume of the construction interpretation result and the original seismic amplitude data;
and training a predetermined initial convolutional neural network model by using the training set to obtain a trained convolutional neural network model so as to predict the geological structure to be predicted by using the trained convolutional neural network model.
According to the intelligent interpretation method for the geological structure, provided by the invention, the label data volume is generated by using the pre-interpretation result of the original seismic amplitude data and by using a method of assigning values to different types of structures and is used as a training set of a convolutional neural network to construct a convolutional neural network model to predict the geological structure. The mutual relation between the horizon and different types of structures is considered, so that the horizon interpretation at the structure is clearer, and the geological interpretation of different structures is more accurate.
Further, the original seismic amplitude data are actual seismic data of a coal field with various typical geological structure types and data of accurate and pre-manual interpretation results; the manual interpretation result comprises: horizon, structure and background.
Further, the performing different types of structural assignment on the three-dimensional data volume of the interpretation result refers to: assigning the label value of the layer position as 1, assigning the label value of the structure as 2, assigning the label values of other non-structure positions of the background as 0, and generating the label data body; the background refers to the absence of a structure.
Further, the training set includes: the tag data volume and the raw seismic amplitude data.
Further, the convolutional neural network model is a three-dimensional convolutional neural network model.
Further, the constructing the convolutional neural network model includes: selecting an initialization parameter; training the convolutional neural network model using the raw seismic amplitude data and the tag data volume; minimizing the loss function results in the final training model.
Further, the convolutional neural network comprises three convolutional layers for feature extraction;
the convolution kernel size of the first convolution layer is 3 multiplied by 3, and the step length is 2; the convolution kernel size of the second convolution layer and the third convolution layer is 3 multiplied by 3, and the step length is 1;
the three convolutional layers each apply a corrective linear activation.
Further, the convolutional neural network also comprises two pooling layers, and features extracted by the convolutional layers are selected and filtered;
the first pooling layer has a size of 2 × 2 × 2 and a step length of 2, and is disposed after the first convolution layer;
the second pooling layer has a size of 2 × 2 × 2 and a step size of 2, and is disposed after the second convolution layer.
Further, the convolutional neural network also comprises a full connection layer and a softmax classifier;
the full-connection layer is activated by adopting linear correction and is arranged behind the third convolution layer;
the softmax classifier is disposed behind the fully-connected layer for generating training and prediction results.
In a second aspect, there is provided a machine-readable storage medium having stored thereon machine-executable instructions which, when invoked and executed by a processor, cause the processor to carry out the method of the first aspect.
According to the intelligent interpretation method of the geological structure, the manual pre-interpretation result of the actual seismic data of the coal field is utilized, the label data volume is generated by assigning values to different types of structures and is used as a training set of the convolutional neural network, and a three-dimensional convolutional neural network model is constructed to predict the geological structure. The trained convolutional neural network model is more suitable for actual data of the coal field, and the mutual relation between the horizon and different types of structures is considered, so that the horizon interpretation at the structure is more accurate; the trained convolutional neural network model can be directly used for predicting any other coal field data, and can automatically explain the geological structure quickly and accurately.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of a geological structure interpretation method provided by an embodiment of the invention;
FIG. 2 is a training set of a geological structure interpretation method provided by an embodiment of the present invention;
FIG. 3 is a convolutional neural network structure of a geological structure interpretation method according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating the actual coal field geological structure prediction results of a geological structure interpretation method according to an embodiment of the present invention;
fig. 5 is a flow chart of another geologic structure interpretation method provided by the embodiment of the invention.
Detailed Description
To make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some, but not all embodiments of the present invention. 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.
With the continuous development of coal resource exploration, the accuracy requirement of seismic data interpretation is higher and higher. Initially, horizon interpretation and structure interpretation were performed by manual interpretation, which was time-consuming and lacking in objectivity for the interpreter to interpret based on his own geological knowledge and experience.
Aiming at the defects existing in manual interpretation, the seismic fault interpretation is realized by the following two ways:
the first method comprises the following steps: and identifying the fault plane by using the full convolution neural network. Seismic amplitude fault slice data are obtained from the three-dimensional seismic amplitude data volume, fault positions on the slice data are calibrated through a manual interpretation method, and the slice data are used as tag data. The method uses the slice data as labels, does not consider the mutual relation between adjacent slices, and results in inaccurate and unrepresentative prediction results.
And the second method comprises the following steps: and performing model training by using the coherent body attribute as a label and combining the seismic data body, and predicting the fault of the new seismic data body by using the trained network model. The quality of the prediction result of the method depends on the precision of a training model, the training model uses a coherent body as a label, and the classification result is not accurate.
In summary, for the problem of seismic fault interpretation at present, the manual interpretation method is time-consuming and subjective, and the accuracy of the automatic prediction method is low. For the understanding of the present embodiment, the following detailed description will be given of the embodiment of the present invention.
The embodiment provides an intelligent explanation method of geological structure, the flow chart of the method is shown in fig. 1, and the method comprises the following steps:
s110, generating a three-dimensional data body of an interpretation result according to the predetermined interpretation result of the original seismic amplitude data;
s120, determining the type of the interpretation result based on the interpretation result;
s130, assigning a value to the three-dimensional data body of the interpretation result based on the type of the interpretation result, and generating a label data body of the interpretation result;
s140, establishing a training set according to the label data volume of the interpretation result and the original seismic amplitude data;
s150, training the predetermined initial convolutional neural network model by using the training set to obtain the trained convolutional neural network model so as to predict the geological structure to be predicted by using the trained convolutional neural network model.
For step S110, raw seismic amplitude data, which may refer to actual seismic data of a coal field having a plurality of typical geological structure types, may be predetermined, as shown in section (a) of FIG. 2. Next, the raw seismic amplitude data may be manually interpreted to obtain interpretation results, which may include horizons, structures, and so on; and carrying out three-dimensional geological structure modeling on the interpretation result to generate a three-dimensional data volume.
For step S120, the types of interpretation results may include a horizon type, a structure type, a non-structure type, and the like.
For step S130, an assignment may be performed on the three-dimensional data volume according to different types of interpretation results, and the assignment may be used as a label of the three-dimensional data volume. For example, a tag with the value of 1 may be added to the three-dimensional data volume of the horizon type, a tag with the value of 2 may be added to the three-dimensional data volume of the structure type, and a tag with the value of 0 may be added to the three-dimensional data volume of the background type to generate a tag data volume; as shown in part (b) of fig. 2, the gray layer represents a horizon, the black dots represent structures, and the dark gray three sides represent a background, i.e., where there are no structures.
For step S140, a training set of neural network models may be established based on the raw seismic data shown in part (a) of fig. 2 and the tagged data volume shown in part (b) of fig. 2.
For step S150, an initial convolutional neural network model is first established, and relevant parameters are initialized. For example, in some embodiments, the initial three-dimensional convolutional neural network structure is shown in fig. 3, and includes three convolutional layers, two pooling layers, one fully-connected layer, and one softmax classifier.
The convolutional layer is composed of a plurality of characteristic graphs, the convolutional layer parameters comprise the size of a convolutional core, step length and filling, the size of the convolutional layer output characteristic graph is determined by the convolutional layer parameters and the convolutional layer parameters are hyper-parameters of the convolutional neural network. Where the convolution kernel size can be specified as an arbitrary value smaller than the input image size, the larger the convolution kernel, the more complex the input features that can be extracted. And carrying out convolution operation on the convolution kernel and the input image of the previous layer to obtain an image with a smaller size. As shown in formula 1, each element in the convolution kernel is a weight parameter, and the convolution kernels are multiplied and added with the pixel values in the corresponding range in the input image of the layer, and the pixel values of the output image are obtained through the Softmax function. The matrix corresponding to the jth characteristic diagram of the kth layer is obtained by convolution weighting of a plurality of characteristic diagrams of the previous layer and operation of an activation function.
Wherein: f is an activation function, NjIs a combination of the input feature maps,
is a feature matrix of the previous layer of images,
are the weights in the convolution kernel matrix and,
is the jth bias matrix for the kth layer.
The activation function is a nonlinear mapping layer in the convolutional neural network, the ReLU function is the most used activation function in the convolutional neural network, when the input is less than 0, the output is 0, and when the input is more than 0, the output is the original value. ReLU function calculation formula:
f(x)＝max(0,x) (2)
in the convolutional neural network structure shown in fig. 3, the convolutional kernel size used by the first convolutional layer is 3 × 3 × 3, and the step size is 2; the convolution kernel size used by the second and third convolution layers is 3 × 3 × 3, and the step size is 1; each convolutional layer consists of a set of learnable filters that have small perceptual learning domains but can extend the depth of input, and is mainly used for feature extraction.
A pooling layer is typically added between each convolutional layer. After the feature extraction is performed on the convolutional layer, the output feature map is transmitted to the pooling layer for feature selection and information filtering. The pooling layer contains a pre-set pooling function whose function is to replace the result of a single point in the feature map with the feature map statistics of its neighboring regions. The step of selecting the pooling area by the pooling layer is the same as the step of scanning the characteristic diagram by the convolution kernel, and the pooling size, the step length and the filling are controlled. Pooling is divided into maximum pooling, average pooling and random pooling. The embodiment adopts maximum pooling, the maximum value of elements in a block is taken as the output of a function, the local maximum value of a feature plane is extracted, and the maximum pooling calculation formula is as follows:
wherein:
is the eigenvalue, u, at position (i, j) obtained by the w-th convolution kernelwThe maximum value of the block is calculated.
In the convolutional neural network structure shown in fig. 3, maximum pooling of 2 × 2 × 2 in size and 2 steps is used after both the first convolutional layer and the second convolutional layer. The role of the pooling layer is to reduce the size of the picture while preserving the picture basic information and to reduce the overall amount of parameters and calculations with pooling operations.
At the end of the convolutional neural network model, one or more fully-connected layers are typically connected, which serve as classifiers in the convolutional neural network, and all neurons in the previous layer are interconnected with neurons in the next layer. In the convolutional neural network structure shown in fig. 3, a fully-connected layer is disposed after the third convolutional layer, and a corrective linear activation (ReLU) is applied after each convolutional layer and the fully-connected layer.
And the full connection layer maps all the obtained distributed characteristic maps to a sample space, and then the full connection layer performs classification calculation through a softmax function and outputs a class label of the maximum probability corresponding to the input image. Calculation formula of Softmax function:
wherein: n is the number of categories.
In the training phase, the convolutional neural network needs to calculate the difference between the current predicted value and the real value at each iteration, and the difference is the current loss value. The Softmax Loss function can calculate the classification Loss in the current iteration:
wherein: and N is the number of samples in each batch, is a label of the current training sample, and is a score of a corresponding category output by the current iteration.
In the convolutional neural network structure shown in fig. 3, the softmax classifier set behind the fully connected layer generates a result indicating the possibility of each classification occurring at the input center point, and fig. 4 is a prediction result of the actual coal field geology using the convolutional neural network model of fig. 3.
In step S150, the training set is used to train the predetermined initial convolutional neural network model, and then the trained convolutional neural network model is obtained. And finally, predicting the classification of the geological structure of the target area by using the obtained convolutional neural network model, thereby realizing automatic seismic interpretation work.
A specific embodiment of one of the geological structure interpretation methods is explained with reference to fig. 5 as follows:
step S101: generating a three-dimensional data volume; and determining an interpretation result according to the original seismic amplitude data based on the pre-interpretation, and generating a three-dimensional data volume of the interpretation result by constructing three-dimensional geological modeling.
Step S102: classifying the interpretation results; the interpretation results are classified into three categories of horizons, structures and backgrounds.
Step S103: establishing a label data body; and based on the type of the interpretation result, assigning values to the three-dimensional data body for interpretation explanation, assigning the label value of the layer position to be 1, assigning the label value of the structure to be 2, and assigning the label values of other non-structure positions of the background to be 0 to generate the label data body for the interpretation result.
Step S104: establishing a training set; and establishing a training set according to the label data body of the interpretation result and the original seismic amplitude data.
Step S105: constructing a neural network model; and constructing a three-position convolution neural network model and initializing relevant parameters.
Step S106: training a neural network model; and training the predetermined initial convolutional neural network model by using the training set data, and obtaining a final three-dimensional convolutional neural network model by minimizing the loss function.
Step S107: predicting a geological structure to be predicted; and predicting the classification of the geological structure of the target area by using the obtained convolutional neural network model, thereby realizing automatic geological structure interpretation work.
According to the method, the label data body is generated by using the manual interpretation result of the coal field actual seismic data, including the horizon, the fault, the collapse column and the like, through a method of constructing geological modeling, so that the trained convolutional neural network model is more suitable for the coal field actual data; structures such as horizons, faults, collapse columns and the like are predicted and automatically interpreted simultaneously, and the mutual relation between the horizons and the structures is considered, so that the horizons interpretation at the structure position is more accurate; the trained convolutional neural network model can be directly used for predicting any other coal field data, and can automatically explain the geological structure quickly and accurately.
In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of description and simplicity of description, but do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present invention, and they should be construed as being included therein. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (9)
1. An intelligent interpretation method of geological formations, characterized by comprising the steps of:
generating a three-dimensional data volume of an interpretation result according to the interpretation result of the predetermined original seismic amplitude data; the interpretation result comprises: the method comprises the steps of (1) horizon, structure and background, wherein the background refers to a non-structure position;
determining a type of the interpretation result based on the interpretation result;
based on the type of the interpretation result, assigning a value to the three-dimensional data volume of the interpretation result, assigning a label value of the horizon to 1, assigning a label value of the structure to 2, assigning a label value of the background to 0, and generating a label data volume of the interpretation result;
establishing a training set according to the label data body of the interpretation result and the original seismic amplitude data;
and training a predetermined initial convolutional neural network model by using the training set to obtain a trained convolutional neural network model so as to predict the geological structure to be predicted by using the trained convolutional neural network model.
2. The intelligent interpretation method of geological structures according to claim 1, characterized in that said raw seismic amplitude data are actual seismic data of coal field with various typical geological structure types, and data of accurate pre-interpretation result.
3. The intelligent interpretation method of geological structures according to claim 1, characterized in that said training set comprises: the tag data volume and the raw seismic amplitude data.
4. The intelligent interpretation method of geological structures according to claim 1, characterized in that said convolutional neural network model is a three-dimensional convolutional neural network model.
5. The intelligent interpretation method of geological structures according to claim 1, characterized in that said predetermined initial convolutional neural network model comprises: selecting an initialization parameter; training the convolutional neural network model using the raw seismic amplitude data and the tag data volume; minimizing the loss function results in the final training model.
6. The intelligent interpretation method of geological structures according to claim 1, characterized in that said convolutional neural network comprises three convolutional layers for feature extraction;
the convolution kernel size of the first convolution layer is 3 multiplied by 3, and the step length is 2; the convolution kernel size of the second convolution layer and the third convolution layer is 3 multiplied by 3, and the step length is 1;
the three convolutional layers each apply a corrective linear activation.
7. The intelligent interpretation method of geological structures according to claim 6, characterized in that said convolutional neural network further comprises two pooling layers, selecting and filtering the features extracted by said convolutional layers;
the first pooling layer has a size of 2 × 2 × 2 and a step length of 2, and is disposed after the first convolution layer;
the second pooling layer has a size of 2 × 2 × 2 and a step size of 2, and is disposed after the second convolution layer.
8. The intelligent interpretation method of geological structures according to claim 6, characterized in that said convolutional neural network further comprises a fully connected layer and a softmax classifier;
the full-connection layer is activated by adopting linear correction and is arranged behind the third convolution layer;
the softmax classifier is disposed behind the fully connected layer for generating results.
9. A machine-readable storage medium having stored thereon machine-executable instructions which, when invoked and executed by a processor, cause the processor to implement the method of any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910777505.1A CN110441820B (en) | 2019-08-21 | 2019-08-21 | Intelligent interpretation method of geological structure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910777505.1A CN110441820B (en) | 2019-08-21 | 2019-08-21 | Intelligent interpretation method of geological structure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110441820A CN110441820A (en) | 2019-11-12 |
| CN110441820B true CN110441820B (en) | 2020-06-16 |
Family
ID=68437061
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910777505.1A Active CN110441820B (en) | 2019-08-21 | 2019-08-21 | Intelligent interpretation method of geological structure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110441820B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111273352A (en) * | 2020-01-15 | 2020-06-12 | 中国煤炭地质总局勘查研究总院 | Intelligent detection method and device for geological structure and electronic equipment |
| CN113296152A (en) * | 2020-02-21 | 2021-08-24 | 中国石油天然气集团有限公司 | Fault detection method and device |
| CN112130200B (en) * | 2020-09-23 | 2021-07-20 | 电子科技大学 | Fault identification method based on grad-CAM attention guidance |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104280771A (en) * | 2014-10-27 | 2015-01-14 | 中国石油集团川庆钻探工程有限公司地球物理勘探公司 | Three-dimensional seismic data waveform semi-supervised clustering method based on EM algorithm |
| WO2016073483A1 (en) * | 2014-11-05 | 2016-05-12 | Shell Oil Company | Systems and methods for multi-dimensional geophysical data visualization |
| EP3724793A4 (en) * | 2017-12-14 | 2021-10-20 | Services Pétroliers Schlumberger | System and method for simulating reservoir models |
| CN109086773B (en) * | 2018-08-29 | 2022-03-04 | 电子科技大学 | Fault plane identification method based on full convolution neural network |
| CN109492775B (en) * | 2018-11-19 | 2020-05-12 | 中国矿业大学(北京) | Geological structure interpretation detection method, geological structure interpretation detection device and readable storage medium |
| CN109597129B (en) * | 2019-01-07 | 2019-12-17 | 中国地质大学（武汉） | fracture-cavity type oil reservoir beaded reflection characteristic identification method based on target detection |
-
2019
- 2019-08-21 CN CN201910777505.1A patent/CN110441820B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110441820A (en) | 2019-11-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109389128B (en) | Automatic extraction method and device for electric imaging logging image characteristics | |
| CN109800736B (en) | Road extraction method based on remote sensing image and deep learning | |
| CN109614985B (en) | Target detection method based on densely connected feature pyramid network | |
| CN109829399B (en) | Vehicle-mounted road scene point cloud automatic classification method based on deep learning | |
| CN110441820B (en) | Intelligent interpretation method of geological structure | |
| CN106295714B (en) | Multi-source remote sensing image fusion method based on deep learning | |
| CN113392775B (en) | Sugarcane seedling automatic identification and counting method based on deep neural network | |
| CN105069468B (en) | Hyperspectral image classification method based on ridge ripple and depth convolutional network | |
| WO2022032329A1 (en) | System and method for image-based remote sensing of crop plants | |
| CN110533631A (en) | SAR image change detection based on the twin network of pyramid pondization | |
| CN108764247B (en) | Dense connection-based deep learning object detection method and device | |
| Ling et al. | Superresolution land cover mapping using spatial regularization | |
| EA004551B1 (en) | Method for seismic facies interpretation using textural analysis and neutral networks | |
| CN112200362B (en) | Landslide prediction method, landslide prediction device, landslide prediction equipment and storage medium | |
| KR20200095336A (en) | Method for optimizing hyperparameters of auto-labeling device which auto-labels training images for use in deep learning network to analyze images with high precision, and optimizing device using the same | |
| CN111951097A (en) | Enterprise credit risk assessment method, device, equipment and storage medium | |
| CN106600595A (en) | Human body characteristic dimension automatic measuring method based on artificial intelligence algorithm | |
| CN112906662B (en) | Method, device and equipment for detecting change of remote sensing image and storage medium | |
| CN106951912B (en) | A kind of method for building up of fruits and vegetables cosmetic variation identification model and recognition methods | |
| CN109614883A (en) | A kind of tight sand crack intelligent identification Method based on convolutional neural networks | |
| CN111402131B (en) | Method for acquiring super-resolution land cover classification map based on deep learning | |
| Makido et al. | Weighting function alternatives for a subpixel allocation model | |
| CN113591608A (en) | High-resolution remote sensing image impervious surface extraction method based on deep learning | |
| CN111273376B (en) | Downscaling sea surface net radiation determination method, system, equipment and storage medium | |
| CN115630577A (en) | Downscaling rainfall prediction method and system |
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 |