Model-driven formation contrast image generation method, device and equipment

Abstract

The application discloses a model-driven stratum contrast image generation method, device and equipment, which is applied to the technical field of oil and gas reservoir development and comprises the following steps: inputting well logging curves of different wells in a to-be-detected stratum into an image segmentation model to obtain stratum prediction images corresponding to the wells output by the image segmentation model; splicing the stratum prediction images of the wells with well logging curves of the corresponding wells to obtain initial stratum contrast images of the corresponding wells; and adding stratum lines in the initial stratum contrast images of the wells to obtain target stratum contrast images of the corresponding wells. The above scheme can effectively improve the reliability and intuitiveness of stratum contrast by splicing the stratum prediction images with the well logging curves and adding the stratum lines.

Description

Technical Field

[0001] This application belongs to the field of oil and gas reservoir development technology, and in particular relates to a model-driven method, apparatus and equipment for generating stratigraphic correlation images. Background Technology

[0002] Stratigraphic correlation of oil and gas reservoirs is one of the key steps in reservoir description and characterization. The accuracy of its results directly determines the construction of the reservoir framework, which in turn controls the spatial distribution of reservoirs within the reservoir, and ultimately has a profound impact on the development of oil and gas reservoirs.

[0003] In traditional stratigraphic correlation methods, geological researchers typically rely on manual division of well logging curves from individual wells to achieve stratigraphic correlation. This method heavily depends on the researchers' expertise and experience, thus exhibiting significant subjectivity and low reliability. Summary of the Invention

[0004] The embodiments of this application provide a model-driven method, apparatus, and device for generating stratigraphic correlation images, which can at least improve the reliability of stratigraphic correlation to a certain extent.

[0005] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part from practice of this application.

[0006] According to a first aspect of the embodiments of this application, a model-driven method for generating stratigraphic correlation images is provided, comprising:

[0007] The logging curves of different wells in the formation to be tested are input into the image segmentation model to obtain the formation prediction image corresponding to each well output by the image segmentation model.

[0008] The formation prediction images of each well are stitched together with the logging curves of the corresponding well to obtain the initial formation comparison images of the corresponding wells;

[0009] Add formation lines to the initial formation contrast images of each well to obtain the target formation contrast images for the corresponding wells.

[0010] In some embodiments, the image segmentation model is trained based on the SegNet model or the U-Net model.

[0011] In some embodiments, the method for generating model-driven stratigraphic correlation images further includes:

[0012] High-frequency components are extracted from the logging curves of the training well within the formation to be tested to obtain high-frequency curves;

[0013] After standardizing the high-frequency curve, it is converted into a grayscale image to obtain the sample image;

[0014] The sample images and stratigraphic label images are input into the SegNet or U-Net model for training until the SegNet or U-Net model converges, thus obtaining the image segmentation model.

[0015] In some embodiments, the U-Net model includes an encoder and a decoder, wherein the encoder includes convolutional layers, batch normalization layers, activation function layers, and pooling layers.

[0016] In some embodiments, the formation prediction images of each well are stitched together with the logging curves of the corresponding wells, including:

[0017] Based on the wellbore depth range corresponding to the formation prediction image of each well and the wellbore depth range of the logging curve of the corresponding well, the formation prediction image and logging curve of the corresponding well are stitched together.

[0018] In some embodiments, formation lines are added to the initial formation contrast images of each well to obtain the target formation contrast image for the corresponding well, including:

[0019] Based on the formation prediction images of each well and the corresponding wellbore depth range, the depth of each target layer in the corresponding well is determined.

[0020] Based on the depth of each target stratum in each well, formation lines are added to the initial formation contrast image of the corresponding well to obtain the target formation contrast image of the corresponding well.

[0021] In some embodiments, formation lines are added to the initial formation contrast images of each well to obtain the target formation contrast image for the corresponding well, including:

[0022] In response to a stratification adjustment command sent by the user based on a stratum prediction image, the stratigraphic lines in the target stratigraphic comparison image are adjusted.

[0023] According to a second aspect of the embodiments of this application, a model-driven stratigraphic correlation image generation apparatus is provided, comprising:

[0024] The layered prediction module is used to input the logging curves of different wells in the formation to be tested into the image segmentation model to obtain the formation prediction image corresponding to each well output by the image segmentation model.

[0025] The image stitching module is used to stitch together the formation prediction images of each well with the logging curves of the corresponding well to obtain the initial formation comparison images of the corresponding wells.

[0026] The formation line addition module is used to add formation lines to the initial formation comparison images of each well to obtain the target formation comparison image of the corresponding well.

[0027] According to a third aspect of the embodiments of this application, a model-driven stratigraphic contrast image generation apparatus is provided, including a processor and a memory, wherein the memory stores computer program instructions that can be executed by the processor, and when the processor executes the computer program instructions, it implements the steps of the method as described in any of the first aspects above.

[0028] According to a fourth aspect of the embodiments of this application, a computer-readable storage medium is provided, which stores computer program instructions that, when executed by a processor, cause the processor to perform the steps of the method as described in any of the first aspects above.

[0029] In this application, well logging curves from different wells within the formation to be tested are input into an image segmentation model to obtain formation prediction images corresponding to each well output by the image segmentation model. The formation prediction images of each well are then stitched together with the corresponding well logging curves to obtain initial formation comparison images for that well. Formation lines are added to the initial formation comparison images of each well to obtain the target formation comparison image for that well. This scheme, by stitching together the formation prediction images and well logging curves and adding formation lines, can effectively improve the reliability and intuitiveness of formation comparison.

[0030] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0031] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. In the drawings:

[0032] Figure 1 A schematic flowchart of a model-driven method for generating stratigraphic correlation images according to some embodiments of this application is shown;

[0033] Figure 2 A schematic diagram showing sample images according to some embodiments of this application is provided;

[0034] Figure 3 A schematic diagram of stratigraphic label images according to some embodiments of this application is shown;

[0035] Figure 4 A schematic diagram showing a comparison of logging curves from two wells according to some embodiments of this application is provided.

[0036] Figure 5 It shows Figure 4A schematic diagram of the initial formation comparison images of the two wells;

[0037] Figure 6 It shows Figure 4 A schematic diagram of the target formation comparison images of the two wells;

[0038] Figure 7 A block diagram of an apparatus for generating model-driven stratigraphic correlation images according to some embodiments of this application is shown;

[0039] Figure 8 A schematic diagram of the structure of a model-driven stratigraphic contrast image generation device according to some embodiments of this application is shown. Detailed Implementation

[0040] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0041] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this application. However, those skilled in the art will recognize that the technical solutions of this application can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this application.

[0042] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0043] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0044] Figure 1 A schematic flowchart illustrating a model-driven method for generating stratigraphic correlation images according to some embodiments of this application is shown. Figure 1As shown, a model-driven method for generating stratigraphic correlation images is provided, which may include the following steps 101 to 103.

[0045] In step 101, the logging curves of different wells in the formation to be tested are input into the image segmentation model to obtain the formation prediction image corresponding to each well output by the image segmentation model.

[0046] The logging curves include one or more of the following: microelectrode logging curves, resistivity logging curves, natural gamma logging curves, and spontaneous potential logging curves. This application does not limit the types of logging curves used.

[0047] In some embodiments, the image segmentation model is trained based on the SegNet model or the U-Net model.

[0048] Understandably, the SegNet model follows an encoder-decoder design pattern. Its encoder section meticulously constructs 13 convolutional layers, interspersed with 5 max-pooling layers, to extract high-level features of the image layer by layer and reduce spatial dimensionality. Correspondingly, the decoder section includes a combination of upsampling layers and convolutional layers, ensuring that the output of each encoder layer is symmetrically processed by a corresponding decoder layer. The decoding process first utilizes the index information from the pooling process to restore the feature map to near its original size. Then, convolutional layers further refine the features, effectively compensating for any details lost during the encoding stage. Finally, the decoder generates a feature map with the exact same size as the input image, where the value of each pixel precisely reflects the probability that the pixel belongs to each target layer.

[0049] The U-Net model is also built on an encoder-decoder framework, but it exhibits unique features in its details. In some embodiments, the U-Net model includes an encoder and a decoder, wherein the encoder includes convolutional layers, batch normalization layers, activation function layers, and pooling layers.

[0050] In the encoder part of U-Net, batch normalization (BN) and activation functions are conventionally introduced after each convolutional layer to enhance the stability and convergence speed of model training. It also includes pooling layers to reduce spatial dimensionality and extract more abstract features. In the decoder part of U-Net, the size of the feature maps is progressively increased through upsampling layers to match the spatial resolution of the original image. Crucially, U-Net introduces skip connections, which directly concatenate low-level feature maps from the encoder with high-level feature maps of corresponding size in the decoder. This design not only promotes feature reuse but also helps the model more accurately locate and reconstruct boundaries when recovering image details, thus significantly improving the accuracy and performance of image segmentation.

[0051] In some embodiments, high-frequency components can be extracted from the logging curves of training wells within the formation to be tested to obtain high-frequency curves; after standardization of the high-frequency curves, they can be converted into grayscale images to obtain sample images; the sample images and formation label images can be input into the SegNet model or U-Net model for training until the SegNet model or U-Net model converges to obtain an image segmentation model.

[0052] In the implementation process, a training well within the formation to be tested can be selected first, and then the logging curves of the training well can be obtained. Using Coif5 wavelets as the wavelet basis, the number of wavelet decomposition layers is determined based on the complexity of the logging signals in the logging curves and the processing requirements. The logging signals are then decomposed, with each layer separating the logging signal into low-frequency and high-frequency components. As the number of decomposition layers increases, the frequency range of the low-frequency components gradually decreases, while the frequency range of the high-frequency components gradually increases, ultimately resulting in multiple low-frequency and multiple high-frequency components. The two high-frequency components with the highest frequency range can be extracted from the logging curves to obtain the high-frequency curve.

[0053] After obtaining the high-frequency curve, invalid signals, missing signals, logically unreasonable or contradictory signals in the high-frequency curve can be removed, and the processed high-frequency curve can be standardized.

[0054] In the standardization process, Z-score standardization can be performed first to convert the data into a standard normal distribution with a mean of 0 and a standard deviation of 1. Then, 0-1 normalization is performed to convert the resulting curve into an image. Finally, the image is processed in the 0-255 pixel space to obtain the sample image.

[0055] Understandably, the stratigraphic label image pre-labels each target layer of the stratigraphy in the image. Figure 2 Schematic diagrams showing sample images according to some embodiments of this application are provided. Figure 3 The diagram illustrates a stratigraphic label image according to some embodiments of this application. Sample images and stratigraphic label images can be used as a dataset. The dataset can be divided according to a certain ratio to obtain a training dataset, a validation dataset, and a test dataset. The SegNet model or U-Net model can be trained using the training dataset, the validation dataset, and the test dataset. After training, an image segmentation model is obtained.

[0056] The image segmentation model obtained by the above scheme avoids the disordered layer phenomenon in the predicted image, thus improving the reliability and accuracy of the predicted image.

[0057] In step 102, the formation prediction image of each well is stitched together with the logging curve of the corresponding well to obtain the initial formation comparison image of the corresponding well.

[0058] Figure 4 A schematic diagram showing a comparison of logging curves from two wells according to some embodiments of this application is illustrated, such as... Figure 4 As shown, the left side shows the spontaneous potential curve (SP) and natural gamma ray curve (GR) of one well (hereinafter referred to as Well 1), and the right side shows the spontaneous potential curve and natural gamma ray curve of another well (hereinafter referred to as Well 2). Both the spontaneous potential curve and the natural gamma ray curve show top lines and bottom lines, which are used to characterize the depth range of the well string.

[0059] Figure 5 It shows Figure 4 A schematic diagram of the initial formation comparison images of the two wells. (See attached image.) Figure 5 As shown, the left side shows the spontaneous potential curve, formation prediction image, and natural gamma curve of well #1, and the right side shows the spontaneous potential curve, formation prediction image, and natural gamma curve of well #2.

[0060] In some embodiments, the formation prediction image and logging curve of a corresponding well can be stitched together based on the well string depth range corresponding to the formation prediction image of each well and the well string depth range of the logging curve of the corresponding well.

[0061] It is understandable that the well string depth range includes the depth of the top and bottom marker layers of each well string. The logging curve of each well includes information related to the well string depth range, and the formation prediction image of each well also includes information related to the well string depth range. When stitching the formation prediction image and the logging image, it is necessary to align the well string depth range corresponding to the formation prediction image with the well string depth range of the logging curve to improve the accuracy of formation correlation.

[0062] In step 103, formation lines are added to the initial formation comparison images of each well to obtain the target formation comparison images of the corresponding wells.

[0063] Figure 6 It shows Figure 4 A schematic diagram of the target formation comparison images from two wells. (See attached image.) Figure 6 As shown, the lines corresponding to S1 to S10 are all stratigraphic lines, and the depth of the target strata increases sequentially from S1 to S10.

[0064] In some embodiments, the depth of each target layer of a corresponding well can be determined based on the formation prediction image of each well and the well string depth range corresponding to the formation prediction image; based on the depth of each target layer of each well, formation lines are added to the initial formation comparison image of the corresponding well to obtain the target formation comparison image of the corresponding well.

[0065] In the implementation process, the image segmentation effect in the formation prediction image and the well depth range corresponding to the formation prediction image can be comprehensively considered to divide the formation into multiple target layers and determine the depth of each target layer. Then, formation lines are generated at the depth of the target layers in the initial formation comparison image to obtain the target formation comparison image of the well.

[0066] In some embodiments, in response to a stratification adjustment instruction sent by a user based on a stratification prediction image, the stratigraphic lines in the target stratigraphic correlation image are adjusted.

[0067] Understandably, considering that the formation prediction image output by the image segmentation model may contain stripes or patches, making it impossible to correctly resolve the location of the target layer in the formation, a manual layer adjustment function can be added. Users can adjust the layer position by referring to the shape of the well logging curve and the formation prediction image. After the user adjusts the layer position, the controller receives the layer adjustment command and adjusts the formation line accordingly.

[0068] By comparing formation prediction images and well logging curves, the reference information for precise layer adjustment by users is expanded. Allowing users to manually adjust the data solves the problem of low model accuracy caused by small sample model training, thereby realizing the intelligent integration of intelligent algorithms and fine reservoir description.

[0069] This application embodiment inputs well logging curves from different wells within the formation to be tested into an image segmentation model to obtain formation prediction images corresponding to each well output by the image segmentation model. The formation prediction images of each well are then stitched together with the corresponding well logging curves to obtain the initial formation comparison images for that well. Formation lines are added to the initial formation comparison images of each well to obtain the target formation comparison images for that well. This solution, by stitching together the formation prediction images and well logging curves and adding formation lines, makes the formation comparison results intuitive, easy to understand, and easy to modify. This effectively improves the reliability and intuitiveness of formation comparison, facilitating rapid assessment of oil and gas resource potential by operational personnel and providing strong data support and decision-making basis for subsequent oil and gas exploration and development plans.

[0070] The following describes an apparatus embodiment of this application, which can be used to execute the model-driven stratigraphic correlation image generation method in the above embodiments of this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the model-driven stratigraphic correlation image generation method described above.

[0071] Figure 7 A block diagram of an apparatus for generating model-driven stratigraphic correlation images according to some embodiments of this application is shown. Figure 7As shown, the model-driven formation contrast image generation apparatus of this application embodiment includes: a layer prediction module 701, used to input the logging curves of different wells in the formation to be tested into an image segmentation model to obtain the formation prediction image corresponding to each well output by the image segmentation model; an image stitching module 702, used to stitch the formation prediction image of each well with the logging curve of the corresponding well to obtain the initial formation contrast image of the corresponding well; and a formation line adding module 703, used to add formation lines to the initial formation contrast image of each well to obtain the target formation contrast image of the corresponding well.

[0072] In some embodiments, based on the aforementioned scheme, the image segmentation model is trained using a SegNet model or a U-Net model.

[0073] In some embodiments, based on the aforementioned scheme, the hierarchical prediction module 701 is further configured to extract high-frequency components from the logging curves of the training wells in the formation to be tested, and obtain high-frequency curves; after standardizing the high-frequency curves, convert them into grayscale images to obtain sample images; input the sample images and formation label images into the SegNet model or U-Net model for training until the SegNet model or U-Net model converges to obtain an image segmentation model.

[0074] In some embodiments, based on the foregoing scheme, the U-Net model includes an encoder and a decoder, wherein the encoder includes a convolutional layer, a batch normalization layer, an activation function layer, and a pooling layer.

[0075] In some embodiments, based on the aforementioned scheme, the image stitching module 702 is further configured to stitch together the formation prediction image and the logging curve of the corresponding well according to the well string depth range corresponding to the formation prediction image of each well and the well string depth range of the logging curve of the corresponding well.

[0076] In some embodiments, based on the foregoing scheme, the formation line addition module 703 is further configured to determine the depth of each target layer of the corresponding well based on the formation prediction image of each well and the well string depth range corresponding to the formation prediction image;

[0077] Based on the depth of each target stratum in each well, formation lines are added to the initial formation contrast image of the corresponding well to obtain the target formation contrast image of the corresponding well.

[0078] In some embodiments, based on the foregoing scheme, the stratigraphic line adding module 703 is further configured to adjust the stratigraphic lines in the target stratigraphic comparison image in response to a stratification adjustment instruction sent by the user based on the stratigraphic prediction image.

[0079] Based on the same inventive concept, embodiments of this application also provide a model-driven stratigraphic correlation image generation device, see reference. Figure 8The diagram shows a schematic of the structure of a model-driven stratigraphic correlation image generation device according to an embodiment of this application. The model-driven stratigraphic correlation image generation device includes one or more memories 804, one or more processors 802, and at least one computer program (computer program instructions) stored in the memory 804 and executable on the processor 802. When the processor 802 executes the computer program, it implements the method described above.

[0080] Among them, Figure 8 In this document, a bus architecture (represented by bus 800) is used. Bus 800 may include any number of interconnected buses and bridges, linking various circuits including one or more processors represented by processor 802 and memory represented by memory 804. Bus 800 may also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. Bus interface 805 provides an interface between bus 800 and receiver 801 and transmitter 803. Receiver 801 and transmitter 803 may be the same element, i.e., a transceiver, providing a unit for communicating with various other devices over a transmission medium. Processor 802 is responsible for managing bus 800 and general processing, while memory 804 can be used to store data used by processor 802 during operation.

[0081] Based on the same inventive concept, embodiments of this application provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, cause the processor to perform the steps of the method described above.

[0082] Based on the same inventive concept, embodiments of this application provide a computer program product, including a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0083] The functions described herein may be implemented in hardware, software executed by a processor, firmware, or any combination thereof. If implemented in software executed by a processor, the functions may be stored as one or more instructions or codes on or transmitted via a computer-readable medium. Other examples and embodiments are within the scope and spirit of this application and the appended claims. For example, due to the nature of software, the functions described above may be implemented using software executed by a processor, hardware, firmware, hardwired, or any combination thereof. Furthermore, the functional units may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit.

[0084] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0085] The units described as separate components may or may not be physically separate. Similarly, the components of the control device may or may not be physical units; they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0086] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing computer program instructions, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0087] The above description is merely an embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A model-driven method for generating stratigraphic correlation images, characterized in that, include: The logging curves of different wells in the formation to be tested are input into the image segmentation model to obtain the formation prediction image corresponding to each well output by the image segmentation model. The formation prediction images of each well are stitched together with the logging curves of the corresponding well to obtain the initial formation comparison images of the corresponding wells; Add formation lines to the initial formation contrast images of each well to obtain the target formation contrast images for the corresponding wells.

2. The method for generating model-driven stratigraphic correlation images according to claim 1, characterized in that, The image segmentation model is trained based on the SegNet or U-Net model.

3. The method for generating model-driven stratigraphic correlation images according to claim 2, characterized in that, Also includes: High-frequency components are extracted from the logging curves of the training well within the formation to be tested to obtain high-frequency curves; The high-frequency curve is standardized and then converted into a grayscale image to obtain a sample image; The sample image and the stratigraphic label image are input into the SegNet model or the U-Net model for training until the SegNet model or the U-Net model converges, thus obtaining the image segmentation model.

4. The method for generating model-driven stratigraphic correlation images according to claim 2, characterized in that, The U-Net model includes an encoder and a decoder, wherein the encoder includes convolutional layers, batch normalization layers, activation function layers, and pooling layers.

5. The method for generating model-driven stratigraphic correlation images according to claim 1, characterized in that, The step of stitching together the formation prediction images of each well with the logging curves of the corresponding well includes: Based on the wellbore depth range corresponding to the formation prediction image of each well and the wellbore depth range of the logging curve of the corresponding well, the formation prediction image and logging curve of the corresponding well are stitched together.

6. The method for generating model-driven stratigraphic correlation images according to claim 1, characterized in that, The step of adding formation lines to the initial formation contrast images of each well to obtain the target formation contrast image for the corresponding well includes: Based on the formation prediction images of each well and the wellbore depth range corresponding to the formation prediction images, the depth of each target layer of the corresponding well is determined; Based on the depth of each target stratum in each well, formation lines are added to the initial formation contrast image of the corresponding well to obtain the target formation contrast image of the corresponding well.

7. The method for generating model-driven stratigraphic correlation images according to claim 6, characterized in that, The step of adding formation lines to the initial formation contrast images of each well to obtain the target formation contrast image for the corresponding well includes: In response to a stratification adjustment command sent by the user based on the stratum prediction image, the stratigraphic lines in the target stratum comparison image are adjusted.

8. A model-driven stratigraphic correlation image generation apparatus, characterized in that, include: The layered prediction module is used to input the logging curves of different wells in the formation to be tested into the image segmentation model to obtain the formation prediction image corresponding to each well output by the image segmentation model. The image stitching module is used to stitch together the formation prediction images of each well with the logging curves of the corresponding well to obtain the initial formation comparison images of the corresponding wells. The formation line addition module is used to add formation lines to the initial formation comparison images of each well to obtain the target formation comparison image of the corresponding well.

9. A model-driven stratigraphic correlation image generation device, comprising a processor and a memory, characterized in that, The memory stores computer program instructions that can be executed by the processor, and when the processor executes the computer program instructions, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions that, when executed by a processor, cause the processor to perform the steps of the method as described in any one of claims 1 to 7.

Images