Method, system and electronic device for geological structure identification based on deep learning

Abstract

The present invention provides a geological structure recognition method, system and electronic device based on deep learning, which relates to the field of geological feature detection and fusion technology, wherein the geological structure recognition method based on deep learning includes: obtaining an initial image of a sample geological section; annotating and calculating feature weights on the initial image to obtain corresponding geological feature information; training a preset deep learning model based on the geological feature information and image feature information in the initial image to generate a target deep learning model; inputting the image to be identified corresponding to the geological section to be identified into the target deep learning model for identification, and outputting the geological structure information of the geological section to be identified. By using the geological feature information and image feature information obtained by annotation and calculation as training data, the deep learning model is trained, thereby improving the model training accuracy, thereby improving the recognition accuracy and comprehensiveness of the geological structure.

Description

Technical Field

[0001] The present invention relates to the field of geological feature detection and fusion technology, and in particular to a geological structure recognition method, system and electronic equipment based on deep learning. Background Art

[0002] Geological structure identification technology is of great value for understanding rock mass structure and evaluating its engineering properties. At present, in the field of geological structure identification, in addition to traditional manual methods, deep learning-based image recognition is the main method, which has a high recognition rate for situations with low geological conditions. Conventional geological structure identification methods usually use the geological body inclination formula and the rock layer thickness calculation formula to obtain relevant parameters for identification. However, the characteristic information obtained by the geological body inclination formula and the rock layer thickness calculation formula has a limited coverage in terms of specific content, and cannot fully capture the complexity of geological structures, nor can it fully reflect the diversity and complexity of geological structures. The geological body inclination formula only considers the inclination angle of the geological body, ignoring other geological features such as crack density and fault distribution; the rock layer thickness calculation formula only focuses on the thickness change of the rock layer, and fails to cover the structural characteristics and material differences inside the rock layer. How to comprehensively and accurately identify and analyze rock mass structures has become a problem that needs to be solved urgently. Summary of the invention

[0003] The purpose of the present invention is to solve the problem that current identification methods are difficult to comprehensively and accurately identify and analyze rock structures. The present invention provides a geological structure identification method, system and electronic device based on deep learning.

[0004] The technical solution of the embodiment of the present application is implemented as follows:

[0005] The first aspect of the embodiment of the present application provides a geological structure identification method based on deep learning, including:

[0006] Acquire an initial image of the sample geological section;

[0007] Performing annotation processing and feature weight calculation on the initial image to obtain corresponding geological feature information;

[0008] Training a preset deep learning model based on the geological feature information and the image feature information in the initial image to generate a target deep learning model;

[0009] The image to be identified corresponding to the geological section to be identified is input into the target deep learning model for identification, and the geological structure information of the geological section to be identified is obtained as output.

[0010] Optionally, after acquiring the initial image of the sample geological section, the method further comprises:

[0011] The initial image is subjected to denoising, cropping and fine-tuning processing.

[0012] Optionally, the labeling process and feature weight calculation of the initial image to obtain corresponding geological feature information includes:

[0013] The target area in the initial image is marked with a rectangular frame; the target area includes faults, folds, fissures, rock layers, intrusive bodies, mineral veins and sedimentary structures.

[0014] Optionally, the labeling process and feature weight calculation of the initial image to obtain corresponding geological feature information further includes obtaining the corresponding geological feature information G using the following formula:

[0015] ;

[0016] in, is the weight coefficient, which is used to balance the influence of different features. is the inclination of the geological body, is the thickness of the rock layer, is the rock density, is a material property, is the complexity of the geological structure, It is geological age.

[0017] The second aspect of the embodiment of the present application provides a geological structure identification system based on deep learning, including: an information acquisition module, a labeling calculation module, a model training module and an information identification module, wherein:

[0018] The information acquisition module is configured to acquire an initial image of the sample geological segment;

[0019] The annotation calculation module is configured to perform annotation processing and feature weight calculation on the initial image to obtain corresponding geological feature information;

[0020] The model training module is configured to train a preset deep learning model based on the geological feature information and the image feature information in the initial image to generate a target deep learning model;

[0021] The information identification module is configured to input the image to be identified corresponding to the geological section to be identified into the target deep learning model for identification, and output the geological structure information of the geological section to be identified.

[0022] Optionally, it also includes an image preprocessing module; the image preprocessing module is configured as follows:

[0023] The initial image is subjected to denoising, cropping and fine-tuning processing.

[0024] Optionally, the annotation calculation module is configured as follows:

[0025] The target area in the initial image is marked with a rectangular frame; the target area includes faults, folds, fissures, rock layers, intrusive bodies, mineral veins and sedimentary structures.

[0026] Optionally, the annotation calculation module is further configured to obtain the corresponding geological feature information G using the following formula:

[0027] ;

[0028] in, is the weight coefficient, which is used to balance the influence of different features. is the inclination of the geological body, is the thickness of the rock layer, is the rock density, is a material property, is the complexity of the geological structure, It is geological age.

[0029] A third aspect of an embodiment of the present application provides an electronic device, comprising a processor and a memory; the memory stores a computer program, wherein the computer program, when executed by the processor, implements the deep learning-based geological structure identification method described in the first aspect.

[0030] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect.

[0031] Compared with the prior art, the technical solution provided by this application has the following beneficial effects:

[0032] The present invention provides a method, device and equipment for identifying geological structures based on deep learning, which obtains the initial image of the sample geological section, annotates the initial image and calculates the feature weight to obtain the corresponding geological feature information, and then trains the preset deep learning model based on the geological feature information and the image feature information in the initial image to generate a target deep learning model, and uses the target deep learning model to obtain the geological structure information of the geological section to be identified. By using the geological feature information and image feature information obtained by annotation and calculation as training data, the deep learning model is trained, which improves the model training accuracy, thereby improving the recognition accuracy and comprehensiveness of the geological structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram of a process flow of a geological structure identification method based on deep learning provided in an embodiment of the present application;

[0034] Figure 2 A schematic diagram of the structure of a geological structure identification system based on deep learning provided in an embodiment of the present application;

[0035] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0036] Below, embodiments of the present application will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present application. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present application. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present application.

[0037] The terms used herein are only for describing specific embodiments and are not intended to limit the present application. The terms "include", "comprising", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0038] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification, and should not be interpreted in an idealized or overly rigid manner.

[0039] Some block diagrams and / or flow charts are shown in the accompanying drawings. It should be understood that some blocks or combinations thereof in the block diagrams and / or flow charts may be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that these instructions, when executed by the processor, may create a device for implementing the functions / operations described in these block diagrams and / or flow charts.

[0040] In some embodiments, see Figure 1 , Figure 1 A schematic diagram of a flow chart of a geological structure identification method based on deep learning provided in an embodiment of the present application; the geological structure identification method based on deep learning provided in an embodiment of the present application includes:

[0041] S110, obtaining an initial image of the sample geological section.

[0042] Here, an initial image of the sample geological section may be collected based on a high-resolution camera, remote sensing equipment, and drone components, etc. The sample geological section may be a section with an environment similar to that of the geological section to be measured.

[0043] In some embodiments, after obtaining an initial image of the sample geological segment at S110, the geological structure recognition method based on deep learning further includes:

[0044] The initial image is denoised, cropped, and fine-tuned.

[0045] Image denoising is to remove the noise from the initial image based on image processing technology to improve the image clarity. Exemplarily, the image processing technology can be mean filtering, median filtering, Gaussian filtering, Fourier transform, wavelet transform and denoising using deep learning models. Image cropping is to crop images for the purpose of geological structure identification, remove irrelevant parts, and focus on geological structure areas. Image fine-tuning can include contrast fine-tuning and brightness fine-tuning; among them, contrast fine-tuning can be to use histogram equalization to increase image contrast. Brightness fine-tuning can be to use Gamma correction to adjust the image brightness through nonlinear transformation.

[0046] S120, annotating the initial image and calculating feature weights to obtain corresponding geological feature information.

[0047] In some embodiments, S120, the initial image is annotated and feature weights are calculated to obtain corresponding geological feature information, including:

[0048] The target areas in the initial image are marked with rectangular frames; the target areas include faults, folds, fissures, rock layers, intrusions, veins and sedimentary structures.

[0049] In this embodiment, LabelImg can be used for labeling, and the image data obtained after denoising is imported into LabelImg, and the imported image is used as the labeling target, and the labeling is performed by drawing a rectangular frame, wherein the labeling includes classification based on image characteristics and image highlighting. Here, the classification includes separate classification of faults, folds, fissures, rock layers, intrusions, veins, and sedimentary structures in the image. Fault annotation, specifically, draw a rectangular box at the fault position, mark its extension range, and assign a "fault" label; fold annotation, specifically, draw a rectangular box in the fold area, mark its wavy structure, and assign a "fold" label; fracture annotation, specifically, draw a rectangular box at the fracture position, mark its fracture range, and assign a "fracture" label; stratum annotation, specifically, draw a rectangular box in the stratum area, mark its bedding characteristics, and assign a "stratum" label; intrusion annotation, specifically, draw a rectangular box at the intrusion position, mark its boundary, and assign a "intrusion" label; vein annotation, specifically, draw a rectangular box at the vein position, mark its filling situation, and assign a "vein" label; sedimentary structure annotation, specifically, draw a rectangular box at the sedimentary structure position, mark its sedimentary characteristics, and assign a "sedimentary structure" label.

[0050] It should be noted that the process of image annotation involves faults, folds, fissures, rock layers, intrusions, veins and sedimentary structures, which can make the rectangular box highlighted. On this basis, the annotated image is subjected to missing check, consistency check and boundary check. Missing check is to compare the actual image with the annotation file to confirm whether there are unannotated geological structure features; consistency check is to check whether the annotations of the same type of geological structure are consistent and whether the annotation range is unified; boundary check is to check whether the annotation box exceeds the image boundary to ensure that all annotations are within the image range.

[0051] In some embodiments, S120, the initial image is annotated and feature weights are calculated to obtain corresponding geological feature information, and further includes obtaining the corresponding geological feature information G using the following formula:

[0052] ;

[0053] in, is the weight coefficient, which is used to balance the influence of different features. is the inclination of the geological body, is the thickness of the rock layer, is the rock density, is a material property, is the complexity of the geological structure, It is geological age.

[0054] In this embodiment, for the inclination angle The sine function uses Used to reflect the change of inclination angle, for thickness The logarithmic transformation is used Smooth thickness changes and avoid zero value problems, for density The square root of Provides a nonlinear characteristic transformation for material properties The sigmoid function uses Mapping material properties to the (0, 1) interval, for the complexity of geological structures The ratio of the inclination angle is used , used to consider the relationship between geological structure complexity and inclination, geological age The ratio of thickness to , which is used to reflect the interaction between geological age and thickness.

[0055] S130, training a preset deep learning model based on the geological feature information and the image feature information in the initial image to generate a target deep learning model.

[0056] In this embodiment, the geological feature information and the image feature information are combined as additional input features of the preset deep learning model. Specifically, a feature matrix is ​​constructed to store the geological feature information G of all images in a feature matrix. In addition, a convolutional neural network is used to extract the image feature information of the initial image, and the geological feature information and the image feature information are combined to form new input feature information. The new input features are input into the fully connected layer or other classification layer of the preset deep learning model for training and prediction.

[0057] In one example, the preset deep learning model includes an input layer, a hidden layer, and an output layer; wherein the input layer can be used to accept a new feature matrix; the hidden layer includes a series of fully connected layers, which are used in conjunction with activation functions for nonlinear transformation; wherein the fully connected layer performs linear transformation on the input features, and the output layer includes a Softmax layer for classification, which outputs the probability distribution of each category. Training the preset deep learning model includes selecting an optimizer, defining a loss function, and an evaluation indicator to complete the training. Among them, the Adam optimizer can be selected, combined with momentum and adaptive learning rate adjustment, and the loss function can use a cross-extraction loss function for multi-classification tasks. On this basis, based on the training results, the hyperparameters of the learning rate and batch size are adjusted to perform model evaluation.

[0058] S140, inputting the image to be identified corresponding to the geological section to be identified into the target deep learning model for identification, and outputting the geological structure information of the geological section to be identified.

[0059] After the target deep learning model is trained, the image to be identified corresponding to the geological section to be identified is input into the target deep learning model to obtain the geological structure information of the geological section to be identified, including faults, folds, cracks, rock formations, intrusions, veins and sedimentary structures of the geological section to be identified.

[0060] The embodiment of the present application obtains the initial image of the sample geological section, annotates the initial image and calculates the feature weight to obtain the corresponding geological feature information, and then trains the preset deep learning model based on the geological feature information and the image feature information in the initial image to generate a target deep learning model, and uses the target deep learning model to obtain the geological structure information of the geological section to be identified. By using the annotated and calculated geological feature information and image feature information as training data, the deep learning model is trained, which improves the model training accuracy, thereby improving the recognition accuracy and comprehensiveness of the geological structure.

[0061] In some embodiments, see Figure 2 , Figure 2A schematic diagram of the structure of a geological structure identification system based on deep learning provided in an embodiment of the present application. The geological structure identification system 200 based on deep learning provided in an embodiment of the present application includes: an information acquisition module 210, a labeling calculation module 220, a model training module 230 and an information identification module 240, wherein:

[0062] An information acquisition module 210 configured to acquire an initial image of a sample geological segment;

[0063] The annotation calculation module 220 is configured to perform annotation processing and feature weight calculation on the initial image to obtain corresponding geological feature information;

[0064] A model training module 230 is configured to train a preset deep learning model based on the geological feature information and the image feature information in the initial image to generate a target deep learning model;

[0065] The information identification module 240 is configured to input the image to be identified corresponding to the geological section to be identified into the target deep learning model for identification, and output the geological structure information of the geological section to be identified.

[0066] In some embodiments, the geological structure recognition system based on deep learning further includes an image preprocessing module; the image preprocessing module is configured as follows:

[0067] The initial image is denoised, cropped, and fine-tuned.

[0068] In some embodiments, the annotation calculation module 220 is configured to:

[0069] The target areas in the initial image are marked with rectangular frames; the target areas include faults, folds, fissures, rock layers, intrusions, veins and sedimentary structures.

[0070] In some embodiments, the annotation calculation module 220 is further configured to obtain the corresponding geological feature information G using the following formula:

[0071] ;

[0072] in, is the weight coefficient, which is used to balance the influence of different features. is the inclination of the geological body, is the thickness of the rock layer, is the rock density, is a material property, is the complexity of the geological structure, It is geological age.

[0073] The deep learning-based geological structure identification system provided in the embodiment of the present application can implement each process in the embodiment corresponding to the above-mentioned deep learning-based geological structure identification method. To avoid repetition, they will not be described here.

[0074] It should be noted that the deep learning-based geological structure identification system provided in the embodiment of the present application and the deep learning-based geological structure identification method provided in the embodiment of the present application are based on the same application concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned deep learning-based geological structure identification method, and the repeated parts will not be repeated.

[0075] In some embodiments, see Figure 3 , Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. An electronic device 300 provided in an embodiment of the present application includes a processor 310 and a memory 320; the memory 320 stores a computer program, wherein the computer program implements the above-mentioned geological structure identification method based on deep learning when executed by the processor.

[0076] Specifically, the processor 310 may include, for example, a general-purpose microprocessor, an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 310 may also include an onboard memory for cache purposes. The processor 310 may be a single processing unit or multiple processing units for executing different actions of the method flow according to the embodiment of the present application.

[0077] The memory 320 may be any medium capable of containing, storing, conveying, propagating or transmitting instructions. For example, the memory 320 may include, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared or semiconductor system, device, component or propagation medium. Specific examples of the memory 320 include: a magnetic storage device, such as a magnetic tape or a hard disk (HDD); an optical storage device, such as a compact disk (CD-ROM); a random access memory (RAM) or a flash memory; and / or a wired / wireless communication link.

[0078] The present application also provides a computer-readable medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned method for identifying geological structures based on deep learning is implemented. The computer-readable medium may be included in the device / apparatus / system described in the above-mentioned embodiment; or it may exist independently without being assembled into the device / apparatus / system. The above-mentioned computer-readable medium carries one or more programs, and when the above-mentioned one or more programs are executed, the method according to the embodiment of the present application is implemented.

[0079] According to an embodiment of the present application, a computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, optical cable, radio frequency signal, etc., or any suitable combination of the above.

[0080] Those skilled in the art will appreciate that the features described in the various embodiments and / or claims of the present application may be combined and / or combined in a variety of ways, even if such combinations or combinations are not explicitly described in the present application. In particular, without departing from the spirit and teachings of the present application, the features described in the various embodiments and / or claims of the present application may be combined and / or combined in a variety of ways. All of these combinations and / or combinations fall within the scope of the present application. Therefore, the scope of the present application should not be limited to the above-described embodiments, but should be determined not only by the appended claims, but also by the equivalents of the appended claims.

Claims

1. A geological structure identification method based on deep learning, characterized in that: include: Acquire an initial image of the sample geological section; The initial image is annotated and feature weights are calculated to obtain corresponding geological feature information; including: obtaining the corresponding geological feature information G using the following formula: ； in, is the weight coefficient, which is used to balance the influence of different features. is the inclination of the geological body, is the thickness of the rock layer, is the rock density, is a material property, is the complexity of the geological structure, is the geological age; Training a preset deep learning model based on the geological feature information and the image feature information in the initial image to generate a target deep learning model; The image to be identified corresponding to the geological section to be identified is input into the target deep learning model for identification, and the geological structure information of the geological section to be identified is obtained as output.

2. The method for identifying geological structures based on deep learning according to claim 1, characterized in that: After acquiring the initial image of the sample geological section, the method further comprises: The initial image is subjected to denoising, cropping and fine-tuning processing.

3. The method for identifying geological structures based on deep learning according to claim 1, characterized in that: The labeling process and feature weight calculation of the initial image to obtain corresponding geological feature information includes: The target area in the initial image is marked with a rectangular frame; the target area includes faults, folds, fissures, rock layers, intrusive bodies, mineral veins and sedimentary structures.

4. A geological structure identification system based on deep learning, characterized in that: include: Information acquisition module, annotation calculation module, model training module and information recognition module, among which, The information acquisition module is configured to acquire an initial image of the sample geological segment; The annotation calculation module is configured to perform annotation processing and feature weight calculation on the initial image, and obtain the corresponding geological feature information G using the following formula: ； in, is the weight coefficient, which is used to balance the influence of different features. is the inclination of the geological body, is the thickness of the rock layer, is the rock density, is a material property, is the complexity of the geological structure, is the geological age; The model training module is configured to train a preset deep learning model based on the geological feature information and the image feature information in the initial image to generate a target deep learning model; The information identification module is configured to input the image to be identified corresponding to the geological section to be identified into the target deep learning model for identification, and output the geological structure information of the geological section to be identified.

5. The geological structure identification system based on deep learning according to claim 4, characterized in that: It also includes an image preprocessing module; the image preprocessing module is configured as follows: The initial image is subjected to denoising, cropping and fine-tuning processing.

6. The geological structure identification system based on deep learning according to claim 4, characterized in that: The annotation calculation module is configured as follows: The target area in the initial image is marked with a rectangular frame; the target area includes faults, folds, fissures, rock layers, intrusive bodies, mineral veins and sedimentary structures.

7. An electronic device comprising a processor and a memory; the memory stores a computer program, wherein: When the computer program is executed by the processor, the computer program implements the deep learning-based geological structure identification method according to any one of claims 1 to 3.

8. A computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the steps of the method according to any one of claims 1 to 3 are implemented.

Images