Rock Images Research Articles

Dynamic Adversarial Adaptation Network with selective pseudo-labels for enhanced unsupervised domain adaptation in rock microscopic image analysis

The critical role of lithology classification in reservoir exploration is increasingly germinating interest in intelligent rock image classification applications. Nonetheless, the efficacy of these classification methods predominantly hinges on the premise that the independent and identically distributed principles underlie the training data. Real-world training data, amassed from disparate geographical regions, tend to manifest non-uniform distributions, which is a factor that muddles the accuracy of contemporary classification strategies. In light of the aforementioned scenario and considering the scarcity of labeled data in the target domain, we suggest a selective pseudo-label-based domain adversarial adaptation. This approach earmarks the generation of an adversarial network that aligns both the conditional and marginal distributions of the source and target domains concurrently during adversarial training. By infusing a dynamic factor, we straddle an adaptable balance between distributions and integrate label information from the target domain using selective pseudo-labeling within the adversarial training phase. Notably, our class-wise pseudo-label technology application determines the pseudo-label threshold dynamically via the selection factor. The thorough evaluation of our proposed approach across four distinct geographic areas encompassing 12 tasks demonstrated that our model produced an average accuracy of 72.2%, signaling a 30% improvement over the source-only model. Importantly, relative to six traditional Unsupervised Domain Adaptation (UDA) algorithms, our approach performed remarkably better, recording an average accuracy surge of a minimum of 3%. Sensitivity assessments also portrayed the robustness of our model in responding to parameter alterations, causing a trivial 1% accuracy fluctuation.

Evaluation method for the intactness of tunnel face surrounding rock based on tunnel face images

The intactness of rock masses is a fundamental parameter in classifying surrounding rocks. Due to limitations imposed by the extent of tunnel face outcrops, the assessment of rock mass intactness necessitates the manual extraction of the positions, orientations and spacing of joints/fissures. To mitigate the labour-intensive nature of this process, in this paper deep learning is employed to develop an integrated method for the automated extraction of joints/fissures and the quantitative analysis of rock mass intactness. We introduce an image preprocessing method based on multiscale histogram equalization to obtain high-contrast, low-noise images. The DeepIntactness model, which incorporates the strategy of curriculum learning to utilize a large number of unlabelled tunnel rock images for model training, is introduced for the extraction of joints/fissures. Following the extraction of joints/fissures, a multiline centre statistic method based on the rock mass block index method is employed to evaluate the intactness of the most vulnerable part of the tunnel face. By applying this approach to an engineering structure, its capacity to automatically extract and quantitatively evaluate the engineering properties of the surrounding rock mass intactness is demonstrated. Hence, this method provides a novel approach to evaluating the tunnel surrounding rock intactness using two-dimensional images. Supplementary material: The I-RBI calculation process of the cases in this paper is available at https://doi.org/10.6084/m9.figshare.c.7154677

A Novel Regional‐Minima Image Segmentation Method for Fluid Transport Simulations in Unresolved Rock Images

AbstractUnresolved digital rock images are often used to avoid high computational costs and limited field of views associated with processing fine‐resolution rock images. However, segmentation of unresolved images using classical methods is suboptimal due to the presence of the partial‐volume effect. Suboptimal segmentations can significantly influence the geometry and effective properties of the reconstructed models. This study reveals that partial‐volume pixels with high pore fractions remain as regional minima in intensity levels in unresolved images. By identifying these regional‐minima pixels, we can effectively extract pore space obscured by the partial‐volume effect. Based on this observation, we propose a novel segmentation method capable of identifying these regional‐minima partial‐volume pixels and converting them to pure pore pixels, thereby binarizing the digital rock images. The method is validated on sandstone and carbonate rock samples. Our method demonstrates a notable improvement in modeled permeability accuracy, surpassing 50% compared to the thresholding method and over 30% compared to the watershed method. Moreover, models segmented by this approach exhibit smaller pore and throat sizes compared to the substantially overestimated results obtained by classical methods. These findings suggest that the regional‐minima segmentation method effectively corrects for the partial‐volume effect and preserves more detailed pore structures. Consequently, it enhances the quality of binarized rock geometries, leading to improved accuracy in fluid‐flow simulations.

Surface wetting characterization in pore-scale multiphase flow simulations: A Ketton carbonate case study

Direct numerical simulations are commonly applied to X-ray computed microtomography images of porous rocks to measure petrophysical properties, such as relative permeability. To complement and/or expand laboratory data with useful simulations, such simulations need to reflect the true wetting condition of reservoir rock. It is therefore important to investigate approaches for assigning surface wetting conditions in pore-scale multiphase flow simulations. Two-phase Lattice-Boltzmann simulations are conducted on a Ketton carbonate image using different effective wetting models and then compared to pore-scale experimental data on the same rock, which is published elsewhere in (Scanziani et al., 2018). Two different wetting models are established based on the aging of the Ketton sample with crude oil at connate water saturation. Model 1 provides uniform wetting on all oil treated surfaces. Model 2 attempts to capture the multiscale nature of carbonate by allowing for spatially varying wetting conditions on oil treated surfaces based on local microporosity. Assessment criteria are based on pore-scale morphological measures, which include oil surface area coverage, oil phase topology, fluid/fluid curvatures, and contact angles. Overall, Model 1 with an assigned contact angle of greater than +10° above the geometrically measured contact angle from the experimental data provided the best results based on our assessment criteria.

An intelligent lithology recognition system for continental shale by using digital coring images and convolutional neural networks

Lithology identification is a critical component of reservoir evaluation and hydrocarbon development. However, the continental shale rocks, characterized by complex lithologies and developed laminations, present new challenges to traditional lithology identification techniques. In this study, we develop a centimeter-level intelligent identification system for continental shale lithology, leveraging digital image preprocessing and deep learning methods. Specifically, we construct a dataset from 400 m of drilling cores, encompassing 24 types of lithology and 152 lithological transitions. The downsampling and oversampling methods are employed to balance the dataset. The digital image denoising and restoration techniques are then utilized to enhance rock image features. Different neural network models are examined to obtain the optimum predictive results. The results indicate that data balancing improves the TOP1 and TOP5 accuracies by 10% and 5%, respectively. Gaussian low-pass filter denoising, in comparison to other methods, increases the TOP1 and TOP5 accuracies by 2%, establishing it as the superior denoising technique for lithology recognition. Considering both computational efficiency and prediction accuracy, EfficientNet emerges as the most effective model for continental shale image recognition. The optimized model attains up to 60% TOP1 and 95% TOP5 across 22 categories in drilling core samples. This study delves into the intelligent identification of macroscopic drilling core images of continental shale, and the findings can offer technical support for the interpretation and evaluation of continental shale reservoirs.

Interactive machine learning for segmenting pores of sandstone in computed tomography images

In the realm of gas field development and CO2/H2 geological storage, the precise characterization of reservoir parameters stands as a critical role. In recent years, digital rock technology has emerged as a cutting-edge tool for examining micro-pore structures, permeability parameter characteristics, and reservoir flow mechanisms. However, there are still certain defects in image segmentation, where low-density pore bodies in low-resolution images make it difficult to clearly distinguish between pores and solid minerals, affecting the accuracy of pore segmentation. This article aims to improve the accuracy of sandstone pore segmentation. We utilized the interactive machine learning segmentation method to segment pore structures and compare the segmented images obtained with the traditional ImageJ grayscale threshold segmentation methods. Furthermore, we evaluated the effectiveness of different segmentation methods quantitatively using the confusion matrix. The research results indicate that the segmentation quality (F1 score) of the machine learning method is higher than the grayscale methods, and the segmentation effectiveness of machine learning method is more stable. To further validate the accuracy of the machine learning method, we simulated the permeability and the results were compared with the experimental results, and the results show that the deviation in permeability segmented through machine learning methods is the smallest. Therefore, our results demonstrate the effective application of the machine learning method in segmenting sandstone pores in the digital rock images.

Efficiently reconstructing high-quality details of 3D digital rocks with super-resolution Transformer

Accurate pore-scale modeling demands high-quality digital rock images, which should possess a broad imaging field of view (FOV) and high resolution to characterize multi-scale rock components. However, achieving both conditions simultaneously is challenging due to hardware constraints. Super-resolution techniques can mitigate this issue by reconstructing high-resolution details from low-resolution images captured with a wide FOV. To reconstruct high-quality 3D digital rocks, we propose a novel Efficient Attention Super-Resolution Transformer (EAST) model. It integrates self-attention and channel attention mechanisms and undergoes structural optimization. Evaluation demonstrates that EAST achieves superior reconstruction quality with a 1.85 × speedup while reducing parameters by 78 % over the advanced model. Additionally, to tailor to the characteristics of digital rocks and the constrained dataset, we employ a hybrid loss function along with two data augmentation techniques. Visualizations reveal that EAST can resist noise and blur interference, highlighting valuable features such as pore edges and textures. Ultimately, we introduce an approach based on self-supervised fine-tuning to enhance the model robustness. Direct flow simulation verifies that the reconstructed results closely align with high-resolution images in terms of physical accuracy. EAST reduces the relative error of the absolute permeability by 18.5 % and 33 % over the Tricubic method on two external samples.

Enhancing digital rock analysis through generative artificial intelligence: Diffusion models

Within the realm of computer vision, the landscape has been significantly reshaped by the abundance of extensive and diverse datasets, leading to remarkable breakthroughs in image processing. These advancements have reverberated across a wide spectrum of applications, catalyzing transformative outcomes. However, in stark contrast, the field of digital rock analysis finds itself grappling with a conspicuous dearth of data, a challenge that casts a formidable shadow over the effective deployment of computer vision techniques for rock image analysis. In response to this pressing issue, this paper presents a pioneering methodology designed to surmount the hurdles posed by data limitation in the realm of digital rock analysis. At the core of this innovative approach lies the fusion of artificially generated digital rock images, created using a state-of-the-art diffusion model, with their authentic counterparts. This fusion is guided by the overarching objective of augmenting the efficacy of various digital rock analysis applications. This integration endeavors to bridge the gap between the limited available data and the substantial demands of the digital rock analysis domain. The practical significance and potential of this integrated approach are vividly demonstrated through a series of concrete implementations. These include, but are by no means limited to, enhancing image quality to facilitate clearer visualization of intricate rock structures and refining the estimation of petrophysical properties with increased accuracy.

Super-resolution reconstruction of 3D digital rocks by deep neural networks

Digital rock technology provides valuable insights into the pore structure and fluid flow properties of geoenergy resources. Artificial intelligence technology in vision and image processing, especially the image super-resolution, has great potential for digital rock reconstruction and resolution enhancement. However, the analyzed core samples are typically sandstones/carbonates in micro-scale resolutions and in two-dimensional (2D) space, whereas the shale rocks in nano-scale resolutions for unconventional resources or three-dimensional (3D) digital cores are rarely investigated. Additionally, previous studies primarily emphasized image quality from a computer vision perspective, with little consideration given to estimating physical properties of digital rocks using super-resolution techniques.This study presents a very deep super-resolution (VDSR) algorithm, specifically designed to generate high-resolution 3D digital rock images, for nano-scale shale matrix and micro-scale hydraulic fractures. We compare both image quality and permeability accuracy between the original high-resolution images and the super-resolution images reconstructed by the proposed method. The results reveal that the reconstructed images using the proposed method closely resemble the actual images, and effectively reduce errors in permeability computations. This study highlights the applicability of the proposed VDSR algorithm in establishing the detailed structures of 3D nano-scale shale matrix and hydraulic fractured rocks, thus advancing super-resolution techniques in digital core analysis for geoenergy resources development.

Drilling rock image segmentation and analysis using segment anything model

Drilling rock image segmentation and analysis using segment anything model

Estimating accurate and precise strain in low-magnitude penetratively deformed rocks: Integrating forward strain modeling and natural data

Estimating grain-scale strain using commonly used strain methods from a low-magnitude penetratively deformed rock is typically fraught with uncertainties as the measured strain is influenced by the pre-deformed grain shapes and grain-scale strain partitioning. Therefore, developing kinematic and mechanical models of orogenic wedge evolution based on such strain data can be misleading without knowing how accurate and precise those data are. To determine accuracy we created five 2-D synthetic rock images and deformed them with known strain under pure shear, simple shear, and general shear. We generated 4160 strain ratio (Rs) and long-axis orientation (φ) data. To determine precision we calculated 2σ (95%) uncertainties in Rs and φ from 102 quartz-rich rocks in Sikkim and generated 977 uncertainty data. Results show that the shape matrix eigenvector and stereographic projection methods record the lowest root mean square error, mean absolute error, and 2σ uncertainty values suggesting they are the most accurate and precise, regardless of strain ratio and deformation types. However, the accuracy and precision of φ deteriorate at Rs < 1.5. For most accurate and precise results, using the above mentioned methods, we recommend a minimum of 100 and 75 markers for Rs < 1.5 and Rs > 1.5, respectively.

Using the traditional microscope for mineral grain orientation determination: A prototype image analysis pipeline for optic-axis mapping (POAM).

This paper reports on the development of an open-source image analysis software 'pipeline' dedicated to petrographic microscopy. Using conventional rock thin sections and images from a standard polarising microscope, the pipeline can classify minerals and subgrains into objects and obtain information about optic-axis orientation. Five metamorphic rocks were chosen to test and illustrate the method. Thin sections were imaged using reflected and cross- and plane-polarised transmitted light. Images were taken at different angles of the polariser and analyser (360° with 10° steps), both with and without the full-lambda plate. The resulting image stacks were analysed with a modular pipeline for optic-axis mapping (POAM). POAM consists of external and internal software packages that register, segment, classify, and interpret the visible light spectra using object-based image analysis (OBIAS). The mapped fields-of-view and grain orientation stereonets of interest are presented in the context of whole-slide images. Two innovations are reported. First, we used hierarchical tree region merging on blended multimodal images to classify individual grains of rock-forming minerals into objects. Second, we assembled a new optical mineralogy algorithm chain that identifies the mineral slow axis orientation. The c-axis orientation results were verified with scanning electron microscopy electron backscattered diffraction (SEM-EBSD) data. For quartz (uniaxial) in a granite mylonite the test yielded excellent correspondence of c-axis azimuth and good agreement for inclination. For orthorhombic orthopyroxene in a deformed garnet harzburgite, POAM produced acceptable results for slow axis azimuth. In addition, the method identified slight anisotropy in garnet that would not be appreciated by traditional microscopy. We propose that our method is ideally suited for two commonly performed tasks in mineralogy. First, for mineral grain classification of entire thin sections scans on blended images to provide automated modal abundance estimates and grain size distribution. Second, for prospective fields of view of interest, POAM can rapidly generate slow axis crystal orientation maps from multiangle image stacks on conventionally prepared thin sections for targeting detailed SEM-EBSD studies.

Controllable image expansion of rock castings based on deep learning

Abstract Digital rock physics (DRP) offers an effective method of deriving elastic parameters from digital rock images, but its practical application is always limited to limited datasets. Recently, deep learning techniques have presented a promising avenue for generating more extensive and cost-effective samples. However, generating controllable samples according to user definition remains very difficult due to high dependence on sufficient datasets. To resolve this problem, a new network was proposed based on the UNet framework through image translation (UNet-IT) to expand rock castings by given porosity in relatively fewer datasets. Practical tests on carbonate rock images demonstrate that the proposed method can generate samples tailored to specific porosity requirements, which achieved a minimum porosity relative error of less than 1%. Compared with the unextended samples, the generated ones have completely different pore structures in terms of two-point probability, two-point cluster, and lineal path functions. Furthermore, the elastic parameters of the generated images obtained through the finite element method (FEM) and practical logging data matched well, with an average relative error of ∼9%. This indicates that the generated samples can be used as effective data to estimate fine rock physics templates and then improve inversion accuracy.

Study on Numerical Simulation of Gas–Water Two-Phase Micro-Seepage Considering Fluid–Solid Coupling in the Cleats of Coal Rocks

The increasing demand for natural gas energy will promote unconventional natural gas, such as coal seam gas and shale gas, to play a key role in future energy development. The mechanical properties of coal seams are weaker compared with conventional natural gas reservoirs. The fluid–solid coupling phenomenon exists widely at the pore scale and macro scale of coal seams, and runs through the whole process of coalbed gas exploitation. The objective of this study is to establish a microscale gas–water flow model for coalbed methane considering fluid structure coupling. Frist, this study used scanning electron microscopy (SEM) to obtain microscopic pore images of coal rocks. Then, we constructed a numerical model to simulate the movement of coalbed methane and water within the scale of coal cleats based on the Navier–Stokes equation, phase field method, and solid mechanics theory. Finally, we analyzed the effects of injection pressure and wettability on the microscopic two-phase seepage characteristics and displacement efficiency of coal. Our research shows that when the injection pressure is increased from 60 kPa to 120 kPa, the displacement completion time is shortened from 1.3 × 10−4 s to 7 × 10−5 s, and the time is doubled, resulting in a final gas saturation of 98%. The contact angle increases from 45° to 120°, and the final gas saturation increases from 0.871 to 0.992, an increase of 12.2%.

RockS2Net: Rock image classification via a spatial localization siamese network

The acquisition of rock property information is at the core of regional geological survey and mineral exploration, but hand-crafted feature-based methods are heavily influenced by human prior knowledge and have limited transferability. End-to-end deep learning techniques, exemplified by convolutional neural networks (CNNs), have attained significant accomplishments in the domain of image classification. However, previous end-to-end CNN-based methods are hard to focus on image critical areas, and they also cannot make full use of global alignment dependency of the rocks. In this paper, rock image classification via a RockS2Net is proposed. The RockS2Net framework adopts a Siamese architecture, characterized by two branches that share parameters, enabling the efficient extraction of both global and local features. This approach facilitates the extraction of global features from entire images and focuses on critical areas to extract local features. The architecture of spatial transformer network (STN) is introduced to transform microscopic images of rock sections to their critical areas. By fusing local features and global features, the properties obtained from microscopic images of rock sections can be more accurately predicted. To test the robust generalizability of the proposed method, the constructed CHN-Rock images dataset is used for experiments and evaluation. Experimental results show that the accuracy of the proposed RockS2Net on the CHN-Rock image dataset is 2–3% higher than that of other rock image classification networks.

A Lithology Recognition Network Based on Attention and Feature Brownian Distance Covariance

In the context of mountain tunnel mining through the drilling and blasting method, the recognition of lithology from palm face images is crucial for the comprehensive analysis of geological conditions and the prevention of geological risks. However, the complexity of the background in the acquired palm face images, coupled with an insufficient data sample size, poses challenges. While the incorporation of deep learning technology has enhanced lithology recognition accuracy, issues persist, including inadequate feature extraction and suboptimal recognition accuracy. To address these challenges, this paper proposes a lithology recognition network integrating attention mechanisms and a feature Brownian distance covariance approach. Drawing inspiration from the Brownian distance covariance concept, a feature Brownian distance covariance module is devised to enhance the network’s attention to rock sample features and improve classification accuracy. Furthermore, an enhanced lightweight Convolutional Block Attention Module is introduced, with upgrades to the multilayer perceptron in the channel attention module. These improvements emphasize attention to lithological features while mitigating interference from background information. The proposed method is evaluated on a palm face image dataset collected in the field. The proposed method was evaluated on a dataset comprising field-collected images of a tunnel rock face. The results illustrate a significant enhancement in the improved model’s ability to recognize rock images, as evidenced by improvements across all objective evaluation metrics. The achieved accuracy rate of 97.60% surpasses that of the current mainstream lithology recognition neural network.

A novel few-shot learning framework for rock images dually driven by data and knowledge

In the field of geosciences, the integration of artificial intelligence is transitioning from perceptual intelligence to cognitive intelligence. The simultaneous utilization of knowledge and data in the geoscience domain is a universally addressed concern. In this paper, based on the interpretability of deep learning models for rock images, rock features such as structure, texture, mineral and macroscopic identification characteristics were selected to extract a rock identification subgraph from the petrographic knowledge graph and carry out rock type similarity reasoning. Comparative experiments were conducted on few-shot learning of rock images under the supervision of rock type similarity knowledge. The results of the few-shot learning comparisons demonstrate that the supervision of rock type similarity knowledge significantly enhances performance. Additionally, rock type similarity knowledge exhibits a marginal effect on improving few-shot learning performance. Given the absence of Chinese word embedding and large-scale Chinese pre-trained language models in the geological domain, graph embedding based on domain-specific knowledge graphs in geosciences can offer computable geoscience knowledge for research dually propelled by data and knowledge.

Machine Learning and Data Fusion Approach for Elastic Rock Properties Estimation and Fracturability Evaluation

Accurate rock elastic property determination is vital for effective hydraulic fracturing, particularly Young's modulus due to its link to rock brittleness. This study integrates interdisciplinary data for better predictions of elastic modulus, combining data mining, experiments, and calibrated synthetics. We used the microstructural insights extracted from rock images for geomechanical facies analysis. Additionally, the petrophysical data and well logs were correlated with shear wave velocity (Vs) and Young's modulus. We developed a machine-learning workflow to predict Young's modulus and assess rock fracturability, considering mineral composition, geomechanics, and microstructure. Our findings indicate that artificial neural networks effectively predict Young's modulus, while K-Means clustering and hierarchical support vector machines excel in identifying rock and geomechanical facies. Utilizing Microscale thin section analysis in conjunction with fracture modeling enhances our understanding of fracture geometries and facilitates fracturability assessment. Notably, fracturability is controlled by specific geomechanical facies during initiation and propagation and influenced by continuity of geomechanical facies in small depth intervals. In conclusion, this study demonstrates data mining and machine learning potential for predicting rock properties and assessing fracturability, aiding hydraulic fracturing design optimization through diverse data and advanced methods.

Digital core: study of textural heterogeneities in a rock

Relevance. When real fluid flows through channels of limited dimensions, it experiences a constant loss of mechanical energy (hydraulic resistance) due to the action of viscous friction. Hydraulic drag has two components: drag along the length of the flow and local drag. The drag along the length of the flow arises because the fluid encounters a force per unit area perpendicular to the direction of the flow. Local drag results from the viscous interaction between the fluid and the channel walls. Local hydraulic resistances include sudden and gradual constrictions/expansions (diffusers/confusions) of the fluid channel, angular and choke gaps. The value of the hydraulic resistance coefficient depends on the type of channel geometry, conditions of fluid entry (wettability of the rock texture) and on the flow regime. When fluid flows through channels of different cross-section, as the channel narrows, the flow velocity increases and the pressure, based on Bernoulli's equation, decreases. Objects. Polymictic sandstones of the Tyumen Formation. Methods. The developed algorithms for identifying areas of contraction/expansion (confusers/diffusers) in the rock texture (polymictic sandstone) are based on neural network reconstruction of core data. The study of geometric diversity of geological rock images is based on generalization of real spatial forms in combination with the axiom of hyperplanes. A methodical approach for finding distribution of inertial exciters (confusers/diffusers) affecting the character of multiphase flow front changes in the rock texture is based on the metrics for estimating possible distributions. Results. The paper introduces the initial stages of core digital transformation data. It was determined that the main problem of geological data digital projection on a large scale is the need to take into account the systems of inertial effects, both in the available volume of the rock and in the extrapolated space of it. In order to adequately represent the character of multiphase fluid flow for different scales, the authors have developed the algorithms for segmentation of geological images of fluid space with further identification of distribution laws of inertial exciters (confusers/diffusers) in the rock texture. For training and testing of the developed algorithms we used the computer tomography images of core material and detailed resolution images of rock sections. The source code of neural network algorithms was written in the Python programming language with the additional use of non-commercial libraries. The algorithms were tested on real data, confirming by experimental studies in the laboratory of digital research in oil and gas as part of the technological project "Digital Core" (Industrial University of Tyumen, Tyumen), the laboratory of the V.I. Shpilman Research and Analytical Center for Rational Subsoil Use (Khanty-Mansiysk), Core Research Laboratory (University of Tyumen, Tyumen).

Ultrahigh-Resolution Reconstruction of Shale Digital Rocks from FIB-SEM Images Using Deep Learning

Summary Accurate characterization of shale pore structures is of paramount importance in elucidating the distribution and migration mechanisms of fluids within shale rocks. However, the acquisition of high-resolution (HR) images of shale rocks is limited by the precision of the scanning equipment. Even with higher-precision devices, compromising the image field of view becomes inevitable, making it challenging to faithfully represent the actual conditions of shale. We propose a stepwise 3D super-resolution (SR) reconstruction method for shale digital rocks based on the widely used focused-ion-beam scanning electron microscope (FIB-SEM) technique. This method effectively addresses the issues of inconsistent horizontal and vertical resolutions as well as low 3D image resolution in FIB-SEM images. By adopting this approach, we significantly enhance image details and clarity, enabling successful observations of pores smaller than 10 nm within shale and laying a foundation for further pore-scale flow simulations. Furthermore, we extract the pore network model (PNM) from the SR reconstructed digital rock to analyze the pore size distribution, coordination number, and pore-throat ratio of shale samples from the Jiyang Depression. The results demonstrate a pore radius distribution in the range of 0 nm to 40 nm, which aligns with the results from nitrogen adsorption experiments. Notably, pores with radii smaller than 10 nm account for 50% of the total connected pores. The proportion of isolated pores in the SR reconstructed shale PNM is significantly reduced, with the coordination number mainly distributed between 1 and 4. The pore-throat ratio of shale ranges from 1 to 3, indicating a relatively uniform development of pores and throats. This study introduces a novel method for accurately characterizing the shale pore structure, which aids researchers in evaluating the pore size distribution and connectivity of shales.