Method and device for generating seismic data training samples

Abstract

The embodiments of the present disclosure disclose a method and device for generating seismic data training samples. The specific implementation of the training sample generation method includes: obtaining a low signal-to-noise ratio area image set, a low signal-to-noise ratio interference image set and a seismic data image set; synthesizing the low signal-to-noise ratio area interference image with the low signal-to-noise ratio area image in the above image set; synthesizing the low signal-to-noise ratio area interference image obtained in the previous step and the seismic data image in the seismic data image set into a sample image; using the sample image and the low signal-to-noise ratio area image to form a training sample, and processing all the images in the image set to obtain a training sample set. This implementation realizes the diversification of training samples of seismic data in the application scenario of identifying low signal-to-noise ratio areas, and makes training samples easier to obtain.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of seismic data processing and interpretation, and in particular to a method and device for generating seismic data training samples. Background Art

[0002] In the field of seismic data processing, signal-to-noise ratio attributes are often used to evaluate processing effects and discover special geological phenomena in seismic profiles. Most of the existing signal-to-noise ratio attribute extraction methods require the given extraction location and analysis time window, and this information is mainly determined by experienced personnel after preliminary analysis. Manual analysis is not only low in precision but also takes a long time and manpower, and the existing methods cannot divide the boundary range of low signal-to-noise ratio areas. Using convolutional neural networks to segment different signal-to-noise ratio areas of seismic data can convert the problem of identifying the signal-to-noise ratio distribution range into an image segmentation problem. With the continuous development of image segmentation algorithms, more requirements are put forward for the training samples of network models, such as the number of samples and the generalization of samples. Different training samples are required to meet the requirements of training accuracy and actual data application. Therefore, theoretical models with different shapes and signal-to-noise ratios are used as samples for network training, and the trained models are used to identify the boundaries of different signal-to-noise ratio areas in seismic data. Practical applications show that this method has strong generalization ability, which is significantly improved in signal-to-noise ratio recognition and partitioning accuracy compared with traditional methods, greatly reducing the workload of manual analysis. Summary of the invention

[0003] The content of this disclosure is used to introduce concepts in a brief form, and these concepts will be described in detail in the specific implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution for which protection is sought, nor is it intended to limit the scope of the technical solution for which protection is sought. Some embodiments of the present disclosure propose a training sample generation method and device to solve the technical problems mentioned in the above background technology.

[0004] In a first aspect, some embodiments of the present disclosure provide a method for generating seismic data training samples, the method comprising: acquiring a low signal-to-noise ratio area image set, a low signal-to-noise ratio interference image set and a seismic data image set; combining the low signal-to-noise ratio area images and the low signal-to-noise ratio interference images in the above image sets into low signal-to-noise ratio area interference images; combining the low signal-to-noise ratio area interference images obtained in the previous step and the seismic data images in the seismic data image set into sample images; forming training samples based on the sample images and the low signal-to-noise ratio area images, and obtaining a training sample set after processing all the images in the image set.

[0005] In a second aspect, some embodiments of the present disclosure provide a seismic data training sample generation device, the device comprising: an image set acquisition unit, a regional interference generation unit, a sample image generation unit and a sample set generation unit. The image set acquisition unit is configured to acquire a low signal-to-noise ratio regional image set, a low signal-to-noise ratio interference image set and a seismic data image set; the regional interference generation unit is configured to combine the low signal-to-noise ratio regional images and the low signal-to-noise ratio interference images in the above image set into a low signal-to-noise ratio regional interference image; the sample image generation unit is configured to combine the low signal-to-noise ratio regional interference images obtained in the previous step and the seismic data images in the seismic data image set into a sample image; the sample set generation unit is configured to form training samples based on sample images and low signal-to-noise ratio regional images, and obtain a training sample set after processing all the images in the image set.

[0006] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in the first aspect.

[0007] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program implements the method described in the first aspect when executed by a processor. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0009] Figure 1 is a schematic diagram of an application scenario of the training sample generation method disclosed in the present invention;

[0010] Figure 2 It is a process of some embodiments of the training sample generation method according to the present disclosure;

[0011] Figure 3 are some low signal-to-noise ratio area images according to the training sample generation method disclosed in the present invention;

[0012] Figure 4 are some signal-to-noise ratio interference images according to the training sample generation method disclosed in the present invention;

[0013] Figure 5 are some seismic data images according to the training sample generation method disclosed in the present invention;

[0014] Figure 6is a training sample image according to the training sample generation method disclosed herein;

[0015] Figure 7 is a process of an embodiment of the image classification generation method according to the present disclosure;

[0016] Figure 8 is an image segmentation result diagram according to the image classification generation method disclosed in the present invention;

[0017] Fig. 9 is a schematic diagram of the structure of some embodiments of the training sample generating device according to the present disclosure;

[0018] Fig.10 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0020] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0021] Figure 1 It is a schematic diagram of an application scenario of the training sample generation method of some embodiments of the present disclosure.

[0022] exist Figure 1In the application scenario 100, first, the computing device 101 can receive a low signal-to-noise ratio region image set 102, a low signal-to-noise ratio interference image set 103, and a seismic data image set 104. The computing device 101 can extract a low signal-to-noise ratio region image 105 from the low signal-to-noise ratio region image set 102, and the low signal-to-noise ratio region image 105 is also a label of a training sample. The computing device 101 can also read a low signal-to-noise ratio interference image 106 from the low signal-to-noise ratio interference image set 103, and the computing device 101 can also read a seismic image 107 from the seismic data image set 104. Then, the computing device 101 combines the low signal-to-noise ratio region image 105 with the low signal-to-noise ratio interference image 106 into a low signal-to-noise ratio region interference image 108. Then, the computing device 101 combines the seismic data image 107 with the low signal-to-noise ratio region interference image 108 into a sample image 109. Finally, a training sample 110 is obtained based on the sample image 109 and the low signal-to-noise ratio area image 105 as a label. In the schematic diagram 100 of this application scenario, the label is: low signal-to-noise ratio area image. The label value is set to 0 and 1. The value inside the low signal-to-noise ratio area is set to 1, and the value outside the low signal-to-noise ratio area is set to 0. According to the low signal-to-noise ratio area image 105 as the label of the sample image 109, the black rectangle in the low signal-to-noise ratio area image 105 is the low signal-to-noise ratio area with a value of 1, and the values ​​of other ranges are 0. The low signal-to-noise ratio area image 105 is multiplied by the low signal-to-noise ratio interference image 106 to obtain the low signal-to-noise ratio area interference image 108, and the low signal-to-noise ratio area interference image 108 is added to the seismic data image 107 to obtain the sample image 109.

[0023] It should be noted that the computing device 101 can be hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is embodied as software, it can be installed in the hardware devices listed above. It can be implemented as multiple software or software modules for providing distributed services, or as a single software or software module. No specific limitation is made here. Figure 1 The number of computing devices in the embodiment is only illustrative. Any number of computing devices may be provided according to implementation requirements.

[0024] refer to Figure 2 , which shows a process 200 of some embodiments of the training sample generation method according to the present disclosure. The training sample generation method comprises the following steps:

[0025] Step 201: Acquire a low signal-to-noise ratio region image set, a low signal-to-noise ratio interference image set, and a seismic data image set.

[0026] The above-mentioned low signal-to-noise ratio area image set can be a triangle, rectangle, ellipse or other graphics, or a combination of graphics. Figure 3 As shown. The above low signal-to-noise ratio interference image set can be a random interference pattern with different signal-to-noise ratios. As an example Figure 4 As shown. The above seismic data image set can be seismic data of different dip angle formations. As an example Figure 5 shown.

[0027] Step 202: combining the low signal-to-noise ratio region image and the low signal-to-noise ratio interference image in the above image set into a low signal-to-noise ratio region interference image.

[0028] The above-mentioned low signal-to-noise ratio region interference image is obtained by the following steps: the low signal-to-noise ratio region image is used as a label, the value in the region is set to 1, and the value outside the region is set to 0. The low signal-to-noise ratio interference image and the low signal-to-noise ratio region image are multiplied to obtain the low signal-to-noise ratio region interference image.

[0029] Step 203, combining the low signal-to-noise ratio area interference image obtained in step 202 and the seismic data image in the seismic data image set into a sample image.

[0030] The sample image is obtained by adding the low signal-to-noise ratio region interference image and the seismic data image. The label of the above sample image is the low signal-to-noise ratio region image.

[0031] Step 204: compose training samples based on the sample images and the low signal-to-noise ratio area images, and process all the images in the image set to obtain a training sample set.

[0032] Generate training samples based on sample images and their labels, and process all the images in the image set to obtain the training sample set. Figure 6 is a training sample.

[0033] Some embodiments of the present disclosure provide methods that combine each low signal-to-noise ratio region image in the above-mentioned low signal-to-noise ratio region image set with a low signal-to-noise ratio interference image selected from the low signal-to-noise ratio interference image set to form a low signal-to-noise ratio region interference image. The above-mentioned low signal-to-noise ratio region image can be a region image of any shape. This makes the training samples more diverse and makes it easier to obtain training samples in this way.

[0034] Further references Figure 7 , which shows a process 700 of some embodiments of the image classification generation method. The process 700 of the image classification generation method includes the following steps:

[0035] Step 701: Input the seismic data image of the low signal-to-noise ratio area to be identified. Figure 8 (a).

[0036] Step 702: Input the target image into a pre-trained image segmentation network to obtain a low signal-to-noise ratio region segmentation result. The training samples of the classification network are obtained by Figure 2 Or the corresponding method to generate. As an example Figure 8 (b).

[0037] In some embodiments, the pre-trained classification network may be a U-net network or a FCN network.

[0038] Further references Fig. 9 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a seismic data training sample generation device. These device embodiments are similar to Figure 2 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0039] like Fig. 9 As shown, the training sample generation device 900 of some embodiments includes: an image set acquisition unit 901, a regional interference generation unit 902, a sample image generation unit 903 and a sample set generation unit 904. Among them, the image set acquisition unit 901 is configured to acquire a low signal-to-noise ratio regional image set, a low signal-to-noise ratio interference image set and a seismic data image set; the regional interference generation unit 902 is configured to combine the low signal-to-noise ratio regional images and low signal-to-noise ratio interference images in the above image sets into low signal-to-noise ratio regional interference images; the sample image generation unit 903 is configured to combine the low signal-to-noise ratio regional interference images obtained in the previous step and the seismic data images in the seismic data image set into sample images; the sample set generation unit 904 is configured to form training samples based on sample images and low signal-to-noise ratio regional images, and obtain a training sample set after processing all the images in the image set. Refer to the following Fig.10 , which shows a structural schematic diagram of an electronic device 1000 suitable for implementing an embodiment of the present application.

[0040] like Fig.10 As shown, the electronic device 1000 includes a central processing unit (CPU) 1001, which can perform various appropriate operations and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage part 1008 to a random access memory (RAM) 1003. In the RAM 1003, various programs and data required for the operation of the system 1000 are also stored. The CPU 1001, the ROM 1002, and the RAM 1003 are connected to each other via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0041] The following components are connected to the I / O interface 1005: an input section 1006 including a keyboard, a mouse, etc.; an output section 1007 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, a modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I / O interface 1005 as needed. A removable medium 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1010 as needed, so that a computer program read therefrom is installed as needed as the storage section 1008.

[0042] In particular, according to an embodiment of the present invention, the process described with reference to the flowchart above can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer-readable storage medium having a computer program stored thereon, which implements the above-mentioned dynamic tracking method based on the front-end framework when executed by a processor.

[0043] In such an embodiment, the computer program may be downloaded and installed from a network via the communication section 1009 , and / or installed from the removable medium 1011 .

[0044] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0045] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing the present application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0046] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0047] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0048] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0049] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Therefore, these modifications, equivalent substitutions, improvements, etc. made without departing from the spirit of the present invention all fall within the scope of protection claimed in the present invention.

Claims

1. A method for generating seismic data training samples, characterized in that The steps include: (1) Acquire a low signal-to-noise ratio area image set, a low signal-to-noise ratio interference image set, and a seismic data image set; (2) synthesizing the low signal-to-noise ratio region image and the low signal-to-noise ratio interference image in the above image set into a low signal-to-noise ratio region interference image; (3) synthesizing a sample image by combining the low signal-to-noise ratio region interference image obtained in the previous step with the seismic data image in the seismic data set; (4) Sample images and low signal-to-noise ratio area images are used to form training samples, and the training sample set is obtained after processing all the images in the image set.

2. The method according to claim 1, wherein: The low signal-to-noise ratio area image set is obtained by the following steps: each low signal-to-noise ratio area is adjusted in shape to obtain the low signal-to-noise ratio area image set; the shape of the low signal-to-noise ratio area is one of a triangle, a rectangle, and an ellipse, or any combination of the above shapes.

3. The method according to claim 1, wherein: The low signal-to-noise ratio interference image set is obtained by the following steps: using random noise to realize interference sets with different signal-to-noise ratios to obtain the low signal-to-noise ratio interference image set; the above-mentioned low signal-to-noise ratio interference is formed by using random noise in sequence according to the signal-to-noise ratio from 1 to 100, where the signal-to-noise ratio value interval is 10.

4. The method according to claim 1, wherein: The seismic data image set is obtained by the following steps: adjusting the formation dip angle by using the theoretical model seismic data to realize the theoretical model seismic data of different formation dip angles, and obtaining the seismic data image set.

5. The method according to claim 1, wherein: The low signal-to-noise ratio region image and the low signal-to-noise ratio interference image in the above image set are synthesized into a low signal-to-noise ratio region interference image, which is obtained by the following steps: extracting the low signal-to-noise ratio region image and the low signal-to-noise ratio interference image from the image set, using the low signal-to-noise ratio region image as a label, setting the value within the region to 1, and setting the value outside the region to 0; multiplying the low signal-to-noise ratio interference image and the low signal-to-noise ratio region image to obtain the low signal-to-noise ratio region interference image.

6. The method according to claim 1, wherein: The low signal-to-noise ratio area interference image obtained in the previous step and the seismic data image in the seismic data image set are synthesized into a sample image through the following steps: the seismic data image is extracted from the image set, and the low signal-to-noise ratio area interference image and the seismic data image are added together to obtain the sample image.

7. The method according to claim 1, wherein: The sample image and the low signal-to-noise ratio area image are used to form a training sample, and the images in the whole image set are processed to obtain the training sample set, which is obtained by the following steps: the low signal-to-noise ratio area image is used as a label of the sample image, and the training sample is formed by combining the obtained sample image and the low signal-to-noise ratio area image, and the training sample set is obtained by processing the images in the whole image set.

8. A device for generating seismic data training samples, comprising: An image set acquisition unit, a regional interference generation unit, a sample image generation unit and a sample set generation unit, wherein the image set acquisition unit is configured to acquire a low signal-to-noise ratio regional image set, a low signal-to-noise ratio interference image set and a seismic data image set; the regional interference generation unit is configured to combine the low signal-to-noise ratio regional images and the low signal-to-noise ratio interference images in the above image sets into low signal-to-noise ratio regional interference images; the sample image generation unit is configured to combine the low signal-to-noise ratio regional interference images obtained in the previous step and the seismic data images in the seismic data image set into sample images; the sample set generation unit is configured to form training samples based on sample images and low signal-to-noise ratio regional images, and obtain a training sample set after processing all the images in the image set.

9. An electronic device, comprising: one or more processors; A storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7 or the method according to claim 8.

10. A computer readable medium having a computer program stored thereon, wherein: When the program is executed by a processor, the method according to any one of claims 1 to 7 or claim 8 is implemented.

Images