A method for determining in-situ heat treatment conditions for shale based on SEM and image processing techniques.
The method uses deep learning and image processing techniques to enhance SEM images of shale, addressing the challenge of determining optimal heat treatment conditions and enhancing permeability, thereby improving shale oil and gas yield.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- SOUTHWEST PETROLEUM UNIV
- Filing Date
- 2025-11-28
- Publication Date
- 2026-07-29
AI Technical Summary
The heterogeneity of shale reservoirs makes determining optimal heat treatment conditions challenging, leading to inefficient energy use and potential damage from excessive heating, while conventional image processing methods fail to provide clear, high-resolution SEM images for structural analysis.
A method using a deep learning algorithm with a Real-ESRGAN neural network and Trainable Weak Segmentation algorithm to enhance SEM images of shale, allowing for precise determination of heat treatment conditions by quantifying structural changes in porosity and permeability.
Enables high-resolution, detailed image processing of shale SEM images, facilitating the determination of optimal heat treatment conditions that enhance permeability and prevent excessive heating damage, thus improving shale oil and gas yield.
Smart Images
Figure 2026122894000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of heat treatment of shale, and particularly to a method for determining the heat treatment conditions of in-situ shale based on SEM and image processing technology.
Background Art
[0002] Shale oil and gas resources are abundant and considered a next-generation alternative energy source, but development of this type of reservoir has been delayed in many regions, and highly efficient development has yet to be achieved. Shale reservoirs are dense and have low porosity and permeability, making development difficult with conventional methods. In recent years, development of this type of reservoir has been progressing both domestically and internationally using in-situ heat treatment technologies (such as Shell's electric heating technology, Taiyuan University of Technology's steam heating, LLNL's RF heating, and Phoenix Wyoming's microwave heating technology). Under the action of thermal energy, the structure of micropores and microcracks in the shale is expanded, improving the permeability and fluidity of the reservoir, thus increasing the yield of shale oil and gas. The invention described in Chinese patent publication number CN118194614A discloses an evaluation method for improving the permeability of shale reservoirs based on ultra-high temperature heating. This method applies the principle of kaolin and clay structure transformation under high temperatures to shale reservoirs, improving the permeability of the reservoir through the transformation of clay mineral properties and structure at high temperatures of 1200°C, thereby increasing the yield of shale. However, the mechanism by which the pore size of shale expands and the permeability increases under thermal energy is complex, and the extent of pore expansion in shale due to thermal action is not clear. Cui Jingwei et al. studied the expansion of shale samples under pressure conditions and found that the permeability of shale pores clearly improves with increasing temperature, but they found that once the temperature reaches a certain threshold, the permeability of shale pores begins to decrease. From the above, it is clear that the heterogeneity of shale makes its heating threshold conditions (temperature, heating time, etc.) unclear. Furthermore, the heterogeneity of shale also causes non-uniformity during heating, meaning that excessively high temperatures can lead to the waste of a large amount of thermal energy. On the other hand, the degree of change in physical properties such as the permeability of shale pores differs under different heat treatment conditions. If the heat treatment threshold is exceeded, excessive heat to the rock minerals can cause cracks to form, potentially blocking pores and cracks. Therefore, it is necessary to quickly determine an effective and rational method for heat treating shale.
[0003] In recent years, methods combining SEM (Scanning Electron Microscope) and image processing techniques have been applied to the characterization of shale. An invention related to Chinese patent publication number CN116740089A discloses a method for optimizing the division of SEM scan images of rock and soil materials and extracting quantitative parameters. This method extracts microstructure parameters of shale by processing SEM images mainly using MATLAB®. However, conventional image processing methods such as noise reduction using median filters have problems such as blurring of details and contours in the processed image, and have clear limitations. [Overview of the project]
[0004] To overcome the shortcomings of conventional technology, the object of the present invention is to provide a method for determining in-situ heat treatment conditions for shale based on SEM and image processing techniques. This method processes SEM images of rock using a deep learning algorithm and can obtain images with greater detail, clearer contours, and higher resolution compared to conventional image enhancement methods. The invention provides a feasible method for determining in-situ heat treatment means for shale by quantifying the changes in structural parameters of shale after heat treatment using SEM and image processing techniques, determining reasonable heat treatment conditions based on the changes in structural parameters, and having a simple operation method.
[0005] To achieve the above objectives, the means provided by the present invention are as follows.
[0006] A method for determining the in-situ heat treatment conditions of shale based on SEM and image processing techniques is: A step of obtaining shale samples after heat treatment under different conditions using a high-temperature, high-pressure reaction vessel, The steps include: obtaining an SEM image of a shale sample after heat treatment, and performing grayscale processing on the SEM image to obtain a grayscale image; The steps include: obtaining an enhanced image by performing super-resolution enhancement on the grayscale image using an enhanced super-resolution generation counter-network; The steps include extracting the pore structure of the shale sample after heat treatment from the aforementioned enhanced image, The method includes the steps of calculating the porosity of the shale surface of the porous structure after heat treatment under different conditions, and optimizing the heat treatment conditions of the shale by changing the structural parameters of the porosity of the shale surface. Preferably, the different conditions in the heat treatment include different heating temperatures and different heating times.
[0007] Preferably, the method for constructing the enhanced super-resolution generation counter-network is: The steps include collecting SEM images of shale to create a sample dataset, A Real-ESRGAN neural network model is constructed, and the Real-ESRGAN neural network model includes two parts: a generator and a discriminator. The process includes the steps of initializing the generator and training the Real-ESRGAN neural network model using a combination of L1 loss, perceptual loss, and GAN loss to obtain the trained augmented super-resolution generation counter-network.
[0008] Preferably, the number of intraresidual dense blocks in the generator before upsampling is 23, the discriminator uses a U-Net discriminator with spectral normalization to enhance face-to-face learning in image details, the U-Net discriminator has the same size input and output, and the U-Net discriminator outputs pixels that are true values while simultaneously providing pixel information to the generator.
[0009] Preferably, the step of extracting the pore structure of the heat-treated shale sample from the enhanced image is: The steps include: dividing the enhanced image using a trainable weak segmentation image segmentation algorithm to obtain the segmentation result; The process includes the step of iteratively modifying the selected region to correct the division result and thereby obtaining the pore structure of the shale sample after heat treatment. Preferably, the formula for calculating the porosity of the shale surface is: The filename is JPEG2026122894000002.jpg12170.
[0010] The present invention discloses the following technical effects through specific embodiments provided by this invention.
[0011] The method for determining the heat treatment conditions of shale in its original location based on SEM and image processing techniques provided by the present invention is: The present invention includes the steps of: obtaining a shale sample after heat treatment under different conditions using a high-temperature, high-pressure reaction vessel; obtaining an SEM image of the heat-treated shale sample and performing grayscale processing on the SEM image to obtain a grayscale image; performing super-resolution enhancement on the grayscale image using an enhanced super-resolution generation counter-network to obtain an enhanced image; extracting the porosity structure of the heat-treated shale sample from the enhanced image; and calculating the porosity of the shale surface after heat treatment under different temperature conditions and optimizing the heat treatment conditions of the shale based on the change in the porosity of the shale surface. The present invention processes SEM images of rock using a deep learning algorithm and can obtain images with greater detail, clearer contours, and higher resolution compared to conventional image enhancement methods. The present invention provides a feasible method for determining in-situ heat treatment means for shale, which quantifies the changes in structural parameters of shale after heat treatment using SEM and image processing techniques, determines reasonable heat treatment conditions based on the changes in structural parameters, and has a simple operation method. [Brief explanation of the drawing]
[0012] To further clarify embodiments of the present invention or technical means in the prior art, the following will be briefly described using drawings used in the embodiments, and obviously the drawings described below represent only some embodiments of the present invention, and it is possible for those skilled in the art to obtain other drawings based on these drawings, assuming that no creative effort has been put into their work. [Figure 1] This is a flowchart of the method provided by an embodiment of the present invention. [Figure 2]This is a comparison diagram of the original SEM image and the SEM image after Real-ESRGAN enhancement provided by an embodiment of the present invention. [Figure 3] This is a schematic diagram showing the processing results of primitive shale provided by an embodiment of the present invention. [Figure 4] This is a schematic diagram showing the shale results (36h) after treatment at 200°C, as provided by an embodiment of the present invention. [Figure 5] This is a schematic diagram showing the results (36h) of shale after treatment at 400°C, as provided by an embodiment of the present invention. [Figure 6] This is a schematic diagram showing the results (36h) of shale after treatment at 600°C, as provided by an embodiment of the present invention. [Figure 7] This is a schematic diagram showing the shale results (400°C) after 12 hours of treatment, as provided by an embodiment of the present invention. [Figure 8] This is a schematic diagram showing the shale results (400°C) after 24 hours of treatment, as provided by an embodiment of the present invention. [Figure 9] This is a schematic diagram showing the shale results (400°C) after 48 hours of treatment, as provided by an embodiment of the present invention. [Modes for carrying out the invention]
[0013] Hereinafter, the technical solutions according to embodiments of the present invention will be clearly and completely described with reference to the drawings of embodiments of the present invention. Naturally, the embodiments described herein are only a part of the embodiments of the present invention, not all of them. All other embodiments that a person skilled in the art could obtain based on the embodiments of the present invention without any creative work should be included within the scope of the present invention.
[0014] The object of the present invention is to provide a method for determining the heat treatment conditions of shale in-situ based on SEM and image processing technology. By using a deep learning algorithm to process SEM images of rocks, compared with conventional general image enhancement methods, images with more details, clearer contours, and higher resolution can be obtained. Using SEM and image processing technology to quantify the changes in shale structure parameters after heat treatment, based on the changes in structure parameters, reasonable heat treatment conditions are determined, and its operation method is simple, providing a feasible method for determining the heat treatment means of shale in-situ.
[0015] To make the above objects, features, and advantages of the present invention clearer and easier to understand, the present invention will be described in more detail below in conjunction with the drawings and specific embodiments.
[0016] FIG. 1 provides a flowchart of the method provided by an embodiment of the present invention. As shown in FIG. 1, the present invention provides a method for determining the heat treatment conditions of shale in-situ based on SEM and image processing technology. The method includes: Step 100 of obtaining a shale sample after heat treatment under different conditions using a high-temperature and high-pressure autoclave; Step 200 of obtaining an SEM image of the shale sample after heat treatment, performing grayscale processing on the SEM image to obtain a grayscale image; Step 300 of enhancing the resolution of the grayscale image using an enhanced super-resolution generative adversarial network to obtain an enhanced image; Step 400 of extracting the pore structure of the shale sample after heat treatment from the enhanced image; Step 500 of calculating the porosity of the pore structure shale surface after heat treatment under different conditions, and optimizing the heat treatment conditions of the shale through the change in the porosity of the shale surface.
[0017] Specifically, this embodiment has two selection means. 1. Selection of heating temperature: (1) Conduct a heat treatment experiment on shale using a high-temperature and high-pressure autoclave, and heat-treat the shale sample for 36 hours at different temperatures.
[0018] (2) Using a scanning electron microscope, SEM images of the shale after heat treatment are obtained, and then the SEM images of the shale are processed to obtain grayscale images.
[0019] (3) The improved enhanced super-resolution generation counter-network (Real-ESRGAN) enabled super-resolution reconstruction of the image, resulting in a high-resolution image with sharp boundaries. A comparison of the original SEM image with the processed image (Figure 2) revealed that the contours of the pore and crack structures in the processed shale SEM image were even clearer, demonstrating the superior image processing effect. Specifically, it includes the following steps: 1) Create a dataset by collecting SEM images of shale. For the degradation process, a higher-order degradation model, an extension of the first-order degradation model, is adopted and shown in equation (1). In the formula JPEG2026122894000003.jpg9170, D indicates a degradation process, * indicates convolution, ↓ indicates downsampling, n indicates noise, and JPEG refers to an irreversible digital compression technique. 2) A Real-ESRGAN neural network model is constructed, mainly consisting of two parts: a generator and a discriminator. The generator structure is consistent with SRGAN, but the number of Residual-in-Residual Dense Blocks (RRDBs) before upsampling is changed from 16 to 23, significantly improving feature extraction capabilities. Furthermore, each RRDB consists of three residual blocks, and network interpolation is employed for upsampling to further improve learning stability. The discriminator uses a U-Net discriminator with spectral normalization to enhance face-to-face learning in image details. Compared to VCG, U-Net is suitable for learning under complex conditions because the input and output sizes are the same, the output pixels are true values, and pixel information is provided to the generator. 3) The learning phase. Learning is divided into two parts. In the first phase, a generative model with an L1 loss function is used, and in the second phase, after initializing the generator, Real-ESRGAN is trained using a combination of L1 loss, perceptual loss, and GAN loss.
[0020] (4) Using the Trainable Weak Segmentation (TWS) image segmentation algorithm, a skeletal model of the shale is extracted, and the segmentation results are continuously improved by iteratively observing the segmentation results and correcting the selected pore structure regions to obtain a good segmentation effect. Through multiple corrections, regions with grayscale values lower than 40 are segmented into pore structures and crack structures, and regions with grayscale values higher than 40 are segmented into matrix and rock minerals (see Figure 3). Exemplarily, this method employs a point-based whole-range threshold image segmentation algorithm and performs high-speed segmentation of two-dimensional images using a machine learning algorithm. The main steps include step 1) inputting an image and performing segmentation learning by artificially labeling specific features, and step 2) observing the segmentation results and iteratively correcting the selected regions to obtain a good segmentation effect.
[0021] (4) Using Fiji, the pore area in SEM images of shale treated at different temperatures was statistically analyzed, and the porosity of the surface was calculated (Table 1). As shown in Figures 4 to 6, it can be seen that the porosity of the shale surface gradually increases with increasing temperature. However, when the heating temperature reaches 400°C, the rate of increase in surface porosity decreases. Furthermore, SEM images of shale heated at 600°C confirmed that rock minerals were exfoliating between layers due to thermal action. Since the exfoliated rock minerals may be heated and clog pores and cracks, it is considered that a heating temperature of 400-600°C is appropriate.
[0022] JPEG2026122894000004.jpg47170
[0023] 2. Selection of heating time: (1) Using a high-temperature, high-pressure reaction vessel, we conducted heat treatment experiments on shale, performing heat treatment experiments at 400°C for different durations. (2) Using a scanning electron microscope, SEM images of the shale after heat treatment were obtained, and then the obtained SEM images of the shale were processed to obtain grayscale images. (3) Using Real-ESRGAN, enhance SEM images of shale processed with different heating times. (4) By using the Trainable Weak Segmentation (TWS) image segmentation algorithm, a shale skeleton model was extracted, and the segmentation results were iteratively observed and the selected pore structure regions were modified to continuously improve the segmentation results and obtain a good segmentation effect (Figures 7-9). (5) Using Fiji, the pore area in SEM images of shale treated at different temperatures was statistically analyzed, and the surface porosity was calculated (Table 2). As shown in Figures 7 and 9, it can be seen that the surface porosity of the shale gradually increases with increasing heating time. This is because the microporous structure of the shale is sufficiently heated by the extended heating time, thus expanding the pore structure. When the heating time reaches 48 hours, the increase in surface porosity decreases, so it is considered that a heating time of 36 to 48 hours is appropriate.
[0024] JPEG2026122894000005.jpg55170
[0025] The beneficial effects of this invention are as follows.
[0026] (1) By processing rock SEM images using a deep learning algorithm, it is possible to obtain images with more detail, clearer contours, and higher resolution compared to general image enhancement methods.
[0027] (2) Using SEM and image processing techniques, the changes in shale structural parameters after heat treatment are quantified, reasonable heat treatment conditions are determined based on the changes in structural parameters, and the method of operation is simple, providing a feasible method for determining in-situ heat treatment means for shale.
[0028] Each example in this specification will be described sequentially, with emphasis on the differences between each example and the others, and any identical or similar parts between examples may be referenced to one another.
[0029] This specification uses specific examples to illustrate the principles and embodiments of the present invention, but these examples are intended to help understand the methods and core ideas of the present invention. Furthermore, those skilled in the art can make various modifications to the specific embodiments and scope of application based on the ideas of the present invention. As stated above, it should be understood that the contents of this specification are not intended to limit the present invention.
Claims
1. A method for determining the heat treatment conditions of shale in its original location based on SEM and image processing techniques, A step of obtaining shale samples after heat treatment under different conditions using a high-temperature, high-pressure reaction vessel, The steps include: obtaining an SEM image of a shale sample after heat treatment, and performing grayscale processing on the SEM image to obtain a grayscale image; The steps include: obtaining an enhanced image by performing super-resolution enhancement on the grayscale image using an enhanced super-resolution generation counter-network; The steps include extracting the pore structure of the shale sample after heat treatment from the aforementioned enhanced image, A method for determining in-situ heat treatment conditions for shale based on SEM and image processing technology, comprising the steps of: calculating the porosity of the shale surface of the porous structure after heat treatment under different conditions; and optimizing the heat treatment conditions for the shale through the change in the porosity of the shale surface.
2. A method for determining the heat treatment conditions of shale in its original location based on SEM and image processing technology according to claim 1, characterized in that the different conditions in the heat treatment include different heating temperatures and different heating times.
3. The method for constructing the aforementioned enhanced super-resolution generation counter-network is: The steps include collecting SEM images of shale to create a sample dataset, A Real-ESRGAN neural network model is constructed, and the Real-ESRGAN neural network model includes two parts: a generator and a discriminator. A method for determining the heat treatment conditions of a shale raw site based on an SEM and image processing technique according to claim 1, comprising the steps of initializing a generator and training the Real-ESRGAN neural network model using a combination of L1 loss, perceptual loss, and GAN loss to obtain the trained enhanced super-resolution generation counter-network.
4. A method for determining the heat treatment conditions of a shale raw location based on an SEM and image processing technique according to claim 3, characterized in that the number of dense residual blocks within the residual before upsampling of the generator is 23, the discriminator enhances face-to-face learning in image details using a U-Net discriminator equipped with spectral normalization, the U-Net discriminator has the same input and output sizes, and the U-Net discriminator provides pixel information to the generator while the output pixels are true values.
5. The step of extracting the pore structure of the shale sample after heat treatment from the aforementioned enhanced image is as follows: The steps include: dividing the enhanced image using a Trainable Week Segmentation image segmentation algorithm to obtain the segmentation result; A method for determining in-situ heat treatment conditions for shale based on SEM and image processing techniques according to claim 1, comprising the step of iteratively modifying the selected region to correct the division result, thereby obtaining the pore structure of the shale sample after heat treatment.
6. The formula for calculating the porosity of the shale surface is: A method for determining the heat treatment conditions of shale in its original location based on the SEM and image processing technology described in claim 1.