A seismic data recovery method based on Criminisi algorithm

Abstract

The application discloses a seismic data recovery method based on a Criminisi algorithm, and aims to give full play to the role of time-lapse seismic data in predicting oil and gas reservoirs. The classic Criminisi algorithm in image repair is applied to three-dimensional seismic attribute data with a vacancy area, so that the seismic attribute data is completed from the edge of the vacancy part of the pipeline area to the center of the vacancy part, and the completed time-lapse seismic data can provide more complete monitoring data.

Description

Technical Field

[0001] This invention belongs to the field of image restoration technology, specifically relating to a seismic data recovery method based on the Criminisi algorithm. Background Technology

[0002] Because some gas reservoirs are low-permeability and highly heterogeneous, resulting in poor reservoir connectivity and inefficient development, seismic monitoring of the gas reservoirs is necessary to achieve efficient development of the gas field. This monitoring is then used to predict and adjust the remaining gas potential of gas reservoirs that are in the middle or late stages of development or even the final stages of development. Using time-shifted seismic data for remaining gas prediction and potential adjustment can become an important research direction and method for potential adjustment in the D gas field.

[0003] The D gas field possesses 3D drag-and-drop data from 2002 / 2004 and OBN 3D data from 2021-2022, providing a data foundation for conducting time-lapse seismic analysis and residual gas prediction. The gas field has multiple gas groups, all currently in production with wells. Among them, the A gas group exhibits multiple sand bodies in the planar region, with reservoirs predominantly low-permeability and an overall low level of exploitation.

[0004] Time-shifted seismic data for residual gas prediction can be an important research direction and method for tapping and adjusting the potential of old gas fields. Using time-shifted seismic data to predict oil and gas reservoirs requires two sets of data collected at different times in the same work area for seismic monitoring. These two sets of 3D seismic data, collected at different times in the same work area, are 3D towed cable data and 3D OBN data. Due to the presence of pipeline areas on the seabed, geophones must be deployed away from these areas when collecting OBN data. The influence of the pipelines causes data gaps in the 3D OBN data. The attribute data of the missing pipeline sections in the 3D OBN seismic data for this work area contains a large amount of information about the main oil and gas producing areas, significantly impacting the prediction and research of the gas field.

[0005] To fully leverage the role of time-lapse seismic data in predicting oil and gas reservoirs, this invention applies the classic Criminisi algorithm from image inpainting to 3D seismic attribute data with gaps. This allows the seismic attribute data to be filled in from the edge of the gap in the pipeline area toward the center of the gap, providing more complete monitoring data after the time-lapse seismic data is filled in. Summary of the Invention

[0006] In order to solve the technical problems existing in the background art, the present invention aims to provide a seismic data recovery method based on the Criminisi algorithm, which applies the classic Criminisi algorithm in image restoration to three-dimensional seismic attribute data with missing areas, so that the seismic attribute data can be filled from the edge of the missing part of the pipeline area to the center of the missing part.

[0007] To solve the technical problem, the technical solution of the present invention is as follows:

[0008] A seismic data recovery method based on the Criminisi algorithm, the method comprising:

[0009] Step 1: Standardize the 3D seismic attribute data to obtain the standardized 3D seismic attribute data;

[0010] Step 2: Based on the standardized 3D seismic attribute data, determine the repair area and then create patch blocks;

[0011] Step 3: Using the prepared patch block, calculate the priority P(p) of the patch block, and then calculate the patch with the highest priority.

[0012] Step 4: Search for the highest priority patch in the known area. Most similar patch

[0013] Step 5: Find the most similar patch With the highest priority patch Perform the replacement process;

[0014] Step 6: In the highest priority patch Update confidence values ​​within the defined region:

[0015] Step 7: Repeat steps 4, 5, and 6 to repair the entire missing area and complete the recovery of the OBN pipeline area data.

[0016] Furthermore, in step 1, the three-dimensional seismic attribute data is standardized to control the data range within 0-255 of RGB.

[0017] Furthermore, in step 2, the missing areas in the standardized 3D seismic attribute data are set to black, and the known areas are set to white; wherein, the standardized 3D seismic attribute data is the original image layer, and the known areas are the masking layer.

[0018] Furthermore, in step 2, the boundary between the missing data area and the known area is used as the repair edge. A circle with radius R is taken as the center of an attribute value on the repair edge, and this data circle is used as the patch block.

[0019] Furthermore, step 3 specifically includes:

[0020] The priority of a patch block is calculated, and the priority consists of two parts: the confidence function and the decision function. Second, the structure propagation function In the structure propagation function D(p), n p It is to repair the normal vector of point p on the edge. Let be the direction of the isoluminance line at point p, and represent the direction perpendicular to the gradient direction at point p. For grayscale images, the normalization factor a = 255. The priority P(p) = C(p)D(p). A higher confidence level C(p) indicates more known information contained in the entire patch block, while a larger D(p) indicates a more complex linear structure within the patch block. After calculating the patch with the highest priority, let it be...

[0021] Furthermore, step 4 specifically includes:

[0022] Search for the highest priority patch in the known area. The most similar patch, the similarity function d(Ψ) between two generic patches a Ψ b ) is defined as Then the known region and The most similar patch is Ψ q Refers to all patches within a known area, where A is Ψ. q The number of data in I ai and I bi Let d represent the data values ​​at points a and b, and let G represent the gradient vectors at points a and b. The smaller d is, the more similar the two patches are; the larger d is, the less similar the two patches are.

[0023] Compared with the prior art, the advantages of the present invention are as follows:

[0024] This invention provides a seismic data recovery method based on the Criminisi algorithm. After recovering the attribute data of the missing areas in the OBN data pipeline, time-shifted seismic data can be used more effectively to monitor the remaining oil and gas reservoirs. Attached Figure Description

[0025] Figure 1 The main flowchart of a seismic data recovery method based on the Criminisi algorithm of the present invention;

[0026] Figure 2 A diagram showing the priority components of this invention. Detailed Implementation

[0027] The specific implementation of the present invention is described below with reference to embodiments:

[0028] It should be noted that the structures, proportions, sizes, etc. shown in this specification are only used to complement the content disclosed in the specification for those skilled in the art to understand and read, and are not intended to limit the conditions under which the present invention can be implemented. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0029] Furthermore, the terms such as "upper," "lower," "left," "right," "middle," and "one" used in this specification are merely for clarity of description and are not intended to limit the scope of the invention. Any changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of the invention.

[0030] Example 1:

[0031] The Criminisi algorithm is an image inpainting algorithm proposed by Antonio Criminisi et al. of Microsoft Research. This algorithm can automatically fill in missing regions in an image, making it appear intact. Its main idea is to infer the content of the missing regions using known content and structure within the image. Specifically, the Criminisi algorithm uses two steps—marker propagation and texture synthesis—to achieve image inpainting.

[0032] The Criminisi algorithm selects the highest priority pixel p on the edge of the region to be repaired, then constructs an n×n pixel block centered on p, then searches for the most similar sample block in the intact region, and updates the repair information in the template block with the found sample block. Finally, it updates the confidence of the pixels in the repaired block and starts the next iteration of repair until the repair is complete.

[0033] To achieve the goal of restoring the missing 3D seismic attribute data of the pipeline area, such as Figure 1 As shown, this embodiment provides the following technical solution:

[0034] Step 1: Standardization of 3D seismic attribute data. 3D seismic attribute data can be compared to RGB data of digital images. Since RGB values ​​in digital images correspond to colors between 0 and 255, the first step is to standardize the 3D seismic attribute data to control the data range within 0-255.

[0035] Step 2: Preparation of data to be restored and data for the repair area. The Criminisi algorithm requires an original image layer and a masking layer. Here, the image layer is the 3D seismic attribute data after standardization. The missing data in the image layer is set to black (RGB value of 0), and the rest is set to white (RGB value of 255) as the masking layer. This step is to allow the program to determine the repair area.

[0036] Step 3: Create the patch block. The boundary between the missing data area and the known area is called the repair edge. Taking an attribute value on the repair edge as the center, take a circle with radius R, and use this data circle as the patch block.

[0037] Step 4: Calculate the priority of the patch block. For example... Figure 2 As shown, the priority consists of two parts: the confidence function and the confidence function. Second, the structure propagation function In the structure propagation function D(p), n p It is to repair the normal vector of point p on the edge. Let be the direction of the isoluminance line at point p, and represent the direction perpendicular to the gradient direction at point p. For a grayscale image, the normalization factor α = 255. The priority P(p) = C(p)D(p). A higher confidence C(p) indicates more known information contained in the entire patch block, while a larger D(p) indicates a more complex linear structure within the patch block. After calculating the patch with the highest priority, let it be...

[0038] Step 5: Search for the highest priority patch in the known area. The most similar patch. The similarity function d(Ψ) between two generic patches. a Ψ b ) is defined as Then the known region and The most similar patch is Ψ q Refers to all patches within a known area, where A is Ψ. q The number of data in I ai and I bi Let d represent the data values ​​at points a and b, and G represent the gradient vectors at points a and b. The smaller d is, the more similar the two patches are; the larger d is, the less similar the two patches are.

[0039] Step Six: Find the most similar patch With the highest priority patch Replace it.

[0040] Step Seven: In Update confidence values ​​within the defined region:

[0041] Step 8: Repeat steps 5, 6, and 7 to gradually repair the entire missing area and complete the recovery of the OBN pipeline area data.

[0042] In step five, the similarity metric can be replaced with other metrics, such as Euclidean distance, Mahalanobis distance, or absolute distance.

[0043] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0044] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0045] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0046] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

[0047] Many other changes and modifications can be made without departing from the concept and scope of this invention. It should be understood that this invention is not limited to the specific embodiments, and the scope of this invention is defined by the appended claims.

Claims

1. A seismic data recovery method based on the Criminisi algorithm, characterized in that, The method includes: Step 1: Standardize the 3D seismic attribute data to obtain the standardized 3D seismic attribute data; Step 2: Based on the standardized 3D seismic attribute data, determine the repair area and then create patch blocks; Step 3: Calculate the priority of the prepared patch block. Then calculate the highest priority patch. ; Step 4: Search for the highest priority patch in the known area. Most similar patch ; Step 5: Find the most similar patch With the highest priority patch Perform the replacement process; Step 6: In the highest priority patch Update confidence values ​​within the defined region: ; Step 7: Repeat steps 4, 5, and 6 to repair the entire missing area and complete the recovery of the OBN pipeline area data; In step 1, the three-dimensional seismic attribute data is standardized to control the data range within 0-255 of RGB. In step 2, the missing areas in the standardized 3D seismic attribute data are set to black, and the known areas are set to white; wherein, the standardized 3D seismic attribute data is the original image layer, and the known areas are the masking layer. In step 2, the boundary between the missing data area and the known area is used as the repair edge. A circle with radius R is taken as the center of an attribute value on the repair edge, and this data circle is used as the patch block. Step 4 specifically includes: Search for the highest priority patch in the known area. The most similar patch, a similarity function between two generic patches. Defined as + Then the known region and The most similar patch is , Refers to all patches within a known area, where A is... The number of data in the data, and Let d represent the data values ​​at points a and b, and let G represent the gradient vectors at points a and b. The smaller d is, the more similar the two patches are; the larger d is, the less similar the two patches are.

2. The seismic data recovery method based on the Criminisi algorithm according to claim 1, characterized in that, Step 3 specifically includes: The priority of a patch block is calculated, and the priority consists of two parts: the confidence function and the decision function. Second, the structure propagation function. ; in the structure propagation function middle, It is the normal vector of point p on the edge that is being repaired. Let be the direction of the isoluminance lines at point p, and represent the direction perpendicular to the gradient direction at point p. For a grayscale image, the normalization factor is... =255; Priority Confidence level A higher value indicates that the entire patch block contains more known information. The larger the value, the more complex the linear structure within the patch block. After calculating the highest priority patch, set it to... .

Images