Search Results (50)

Accurate petrographic identification from thin-section images is important for geoscientific analysis but remains strongly dependent on expert interpretation. In this study, we organized 2634 polarized-light images from 324 distinct thin sections representing 108 lithologies and paired them with Chinese petrographic descriptions. Four conventional vision models were trained as closed-set lithology–classification baselines, and five CN_CLIP variants were fine-tuned for image–text alignment. To prevent overlap of images from the same thin section across outer folds, thin-section-wise three-fold evaluation was performed after excluding eight lithologies represented by fewer than three distinct thin sections; this protocol therefore comprised 2530 images from 311 thin sections and 100 eligible lithologies. At the fixed 20-epoch endpoint, CN_CLIP-ViT-L/14@336px and CN_CLIP-ViT-H/14 showed bidirectional image–text matching accuracies of 81.39 ± 1.97% and 81.31 ± 0.61%, respectively, whereas the conventional classifiers achieved Top-1 accuracies of 50.37%–58.20% under their separate closed-set task. These metrics correspond to different task formulations and are not directly comparable. Under the complementary image-level 8:1:1 protocol, CN_CLIP-ViT-H/14 achieved image-to-text and text-to-image matching accuracies of 93.75% and 93.38%, respectively, on the hold-out test subset. In an inference-stage ablation using the same checkpoint, hold-out test batches, and evaluation protocol, removing only the explicit lithology name reduced these accuracies to 82.72% and 77.21%, indicating that the rock-name token is informative but not the sole source of cross-modal discrimination. Without task-specific retraining, the same CN_CLIP framework also supported constrained lithology, major-mineral, and polarization-mode matching, with best task-specific accuracies of 90.57%, 66.51%, and 86.42%, respectively. Overall, the results support vision-language learning as a flexible framework for expert-assisted petrographic image–text matching on held-out thin sections within the available dataset, while emphasizing source-aware partitioning and cautious interpretation of task-specific accuracies. Full article

This study investigates CO₂–brine–rock interactions in Entrada Sandstone from the San Juan Basin under reservoir-relevant pressure and temperature conditions. A core plug from 8313 to 8315 ft was exposed to CO₂-saturated synthetic formation brine with a salinity of 16,601 ppm, followed by static aging. Pre- and post-experiment analyses included porosity and permeability measurements, ICP-MS, SEM-EDS, XRD, CT imaging, and petrographic thin-section analysis. Permeability decreased from 4.22 to 1.60 mD, corresponding to a 62.08% reduction, whereas porosity decreased from 12.37% to 11.75%, a change interpreted as within experimental uncertainty. Effluent chemistry showed increases in Ca²⁺, Mg²⁺, K⁺, Si, Sr, and Mn, consistent with carbonate cement dissolution and feldspar/silicate alteration under CO₂-acidified brine conditions. XRD analysis of suspended effluent solids identified quartz, montmorillonite, and illite–montmorillonite mixed-layer clays, indicating fines mobilization. These results suggest that permeability impairment was governed primarily by pore–throat blockage from mobilized clay particles, with possible contribution from secondary carbonate redistribution. The findings emphasize the importance of cement composition and clay mineralogy in evaluating injectivity risks for CO₂ storage in clay-bearing sandstone reservoirs. Full article

This paper proposes an intelligent analysis method of rock sheet based on the segment anything model (SAM) with zero samples, which combines small sample training with deep learning to realize high-precision automatic identification and segmentation of rock minerals, and then converts the segmentation results into vectorized data by using image processing technology to construct the numerical model of rock minerals, and ultimately realizes rock sheet from image identification to physical and mechanical research. The results show that the SAM-based zero-sample segmentation method can accurately and efficiently identify different mineral components in multi-component complex rock flakes. Numerical simulation results show that the numerical model of rock minerals generated by the method can effectively reflect the microstructural characteristics of rocks and accurately predict their mechanical behaviors, and the resulting elastic modulus matches well with the existing literature data, with a relative error of only 3.4%, suggesting that the proposed method provides reasonable predictive capability for rock mechanical behavior. Compared with the traditional measurement methods, this method realizes the automation and intelligence of rock thin-section analysis and enhances the adaptability to different rock samples, providing an efficient tool means for geological exploration, petroleum engineering, and geotechnical research. Full article

Quantitative microstructure evaluation of sedimentary rock thin sections is essential for revealing reservoir flow mechanisms and assessing reservoir quality. However, traditional manual identification is inefficient and prone to subjectivity. Although current deep learning approaches have improved efficiency, most remain confined to single tasks and lack a pathway to translate image recognition into quantifiable geological parameters. Moreover, these methods struggle with cross-scale feature extraction and accurate grain boundary localization in complex textures. To overcome these limitations, this study proposes a three-stage automated analysis framework integrating intelligent lithology identification, sandstone grain segmentation, and quantitative analysis of fabric parameters. To address scale discrepancies in lithology discrimination, Rock-PLionNet integrates a Partial-to-Whole Context Fusion (PWC-Fusion) module and the Lion optimizer, which mitigates cross-scale feature inconsistencies and enables accurate screening of target sandstone samples. Subsequently, to correct boundary deviations caused by low contrast and grain adhesion, the PetroSAM-CRF strategy integrates polarization-aware enhancement with dense conditional random field (DenseCRF)-based probabilistic refinement to extract precise grain contours. Based on these outputs, the framework automatically calculates key fabric parameters, including grain size and roundness. Experiments on 3290 original multi-source thin-section images show that Rock-PLionNet achieves a classification accuracy of 96.57% on the test set. Furthermore, PetroSAM-CRF reduces segmentation bias observed in general-purpose models under complex texture conditions, enabling accurate parameter estimation with a roundness error of 2.83%. Overall, this study presents an intelligent workflow linking microscopic image recognition with quantitative analysis of geological fabric parameters, providing a practical pathway for digital petrographic evaluation in hydrocarbon exploration. Full article

Strike-slip faults and their associated fractures in the ultra-deep marine carbonate reservoirs of the Fuman Oilfield, Tarim Basin, hold significant petroleum geological importance, with the developmental characteristics of fractures being a key factor controlling reservoir productivity. This study targets the F_I17 strike-slip fault zone within the oilfield, where a comprehensive evaluation of fracture effectiveness was performed by integrating geological methods, including core and thin section observation, fluid inclusion thermometry, geophysical fracture identification approaches using imaging logging and seismic data, and geomechanical simulations. The results showed that: (1) structural fractures were developed in at least three stages, predominantly high-angle fractures with their strikes obliquely intersecting the main fault at a small angle, and were affected by multiple episodes of fluid activity, while early-phase fractures exhibited severe filling whereas late-phase fractures had good effectiveness; (2) ultra-deep carbonate rocks contained well-developed stylolites, with extensive horizontal stylolites reducing fracture effectiveness; (3) mechanical effectiveness evaluation parameters were proposed by integrating normal stress, shear stress, and formation pressure, with slip tendency as the dominant indicator, and referenced to the leakage factor and dilation tendency to characterize fracture effectiveness; (4) dynamic effectiveness was assessed using closure/opening pressures, defining a reasonable formation pressure range for hydrocarbon exploitation. The findings of this study can provide theoretical guidance for the further exploration and development of ultra-deep reservoirs in the Fuman Oilfield. Full article

Hydrothermal alteration processes, including silicification, sericitization, carbonatization, chloritization, and epidotization, serve as critical indicators in the exploration of precious metal deposits. The identification of these alterations traditionally relies on expert petrographic analysis of thin sections, a method that is time-intensive and prone to subjective interpretation. Although the automated classification of rock types from thin section images using machine learning (ML) and deep learning (DL) techniques has gained increasing attention, the classification of specific hydrothermal alteration types remains underexplored. This study evaluates the performance of four deep convolutional neural network (CNN) architectures—DenseNet121, ResNet50, VGG16, and InceptionV3—for classifying these five alteration types from thin section images. A new dataset comprising 5000 high-resolution thin section images (1000 per alteration type) was developed and used to train and evaluate the models under four optimization algorithms: Adam, RMSprop, SGD, and Adadelta. Among these, the DenseNet121 model achieved the highest performance, attaining accuracy and F1-score values of 1.00 with both RMSprop and Adam optimizers, while the InceptionV3 model recorded the shortest training time at 662 s. The results demonstrate that CNN-based approaches can effectively automate the classification of hydrothermal alteration types, offering a fast, consistent, and objective alternative to traditional methods. This study highlights the potential of deep learning techniques to enhance geological exploration through the accurate and efficient identification of hydrothermal alteration minerals. Full article

Fine-grained sedimentary rocks exhibit significant textural heterogeneity, often obscured by conventional grain size analysis techniques that require sample disaggregation. We propose a non-destructive, image-based grain size characterization workflow, utilizing stitched polarized thin-section photomicrographs, k-means clustering, and watershed segmentation algorithms. Validation against laser granulometry data indicates strong methodological reliability (absolute errors ranging from −5% to 3%), especially for particle sizes greater than 0.039 mm. The methodology reveals substantial internal heterogeneity within Es3 laminated shale samples from the Shahejie Formation (Bohai Bay Basin), distinctly identifying coarser siliceous laminae (grain size >0.039 mm, Φ < 8 based on Udden-Wentworth classification) indicative of high-energy depositional environments, and finer-grained clay-rich laminae (grain size <0.039 mm, Φ > 8) representing low-energy conditions. Conversely, massive mudstones exhibit comparatively homogeneous grain size distributions. Additionally, a multifractal analysis (Multifractal method) based on the S_50bi/S_50si ratio further quantifies spatial heterogeneity and pore-structure complexity, significantly enhancing facies differentiation and reservoir characterization capabilities. This method significantly improves facies differentiation ability, provides reliable constraints for shale oil reservoir characterization, and has important reference value for the exploration and development of the Bohai Bay Basin and similar petroliferous basins. Full article

The accurate classification of rocks is crucial for applications such as earthquake prediction, resource exploration, and geological analysis. Traditional methods rely on expert examination of thin-section images under a microscope, making the process time-consuming and prone to errors. Recent advancements in deep learning have emerged as a powerful tool for automated rock classification; however, distinguishing between similar rock types such as sedimentary, metamorphic, and magmatic rocks remains a challenge. This study proposes a novel hybrid convolutional neural network (CNN) approach that combines the strengths of VGG16 and EfficientNetV2 architectures for the classification of thin-section rock images. The model, developed using the Feature-Selected Hybrid Network (FSHNet), demonstrates significant improvements over individual models, achieving a 5% increase in accuracy compared to Efficient-NetV2B0 and a 9% increase compared to VGG16. By employing the ReliefF algorithm for feature selection and Support Vector Machines (SVMs) for classification, the model further reduces the dimensionality of the feature space, enhancing computational efficiency. The proposed model has been applied to two different rock datasets. The first dataset consists of 2634 images, categorized into sedimentary, metamorphic, and magmatic rock classes. Additionally, the approach was tested on a second dataset comprising petrographic microfacies images, demonstrating its effectiveness in multiclass geological structure classification. Validation on both datasets shows that the proposed method outperforms popular deep learning models and previous studies, achieving a 3% increase in accuracy. These results highlight that the proposed approach provides a robust and efficient solution for automated rock classification, offering significant advancements for geological research and real-world applications. Full article

An important aspect of geoscience and energy research is the analysis of microscopic images, where the assessment of rock properties combines imaging methods with numerical analysis. Given the significant advancements in generative artificial intelligence technologies in recent years, which have enabled the creation of realistic images, a need arises to assess the authenticity of synthetic visual data compared to authentic geological data images. This article evaluates the potential for identifying artificially generated microscopic rock images. Synthetic images were generated using a widely accessible diffusion model, based on real training data. Expert evaluation noted high realism, though some structural and rock-type differences remained detectable. In the study, image descriptors were analyzed to assess their usefulness in distinguishing synthetic data from real data. Discriminative feature selection was conducted, and the effectiveness of various classification models based on the selected parameter sets was compared. The study also proposes a heuristic coefficient demonstrating discriminative potential for the analyzed images. The results confirm the feasibility of building classifiers for synthetic images that could aid in detecting generated visual data in geological and petrographic research. They also serve as a foundation for further exploration of the importance of individual features in such applications. Full article

Recent exploration efforts in the Qiongdongnan Basin have revealed hydrocarbon resources within granitic basement rocks in buried hill traps. However, the formation mechanisms and primary controlling factors of these reservoirs remain poorly understood. In this study, we utilized data from six wells in the Qiongdongnan Basin, including sidewall cores, thin sections, imaging logging, and seismic reflection profiles, to analyze the petrological characteristics, pore systems, and fracture networks of the deep basement reservoir. The aim of our study was to elucidate the reservoir formation mechanisms and identify the key controlling factors. The results indicate that the basement lithology is predominantly granitoid, intruded during the late Permian to Triassic. These rocks are characterized by high felsic mineral content (exceeding 90% on average), with them possessing favorable brittleness and solubility properties. Fractures identified from sidewall cores and interpreted from image logging can be categorized into two main groups: (1) NE-SW trending conjugate shear fractures with sharp dip angles and (2) NW-SE trending conjugate shear fractures with sharp angles. An integrated analysis of regional tectonic stress fields suggests that the NE-trending fractures and associated faults were formed by compressional stresses related to the Indosinian closure of the ancient Tethys Ocean. In contrast, the NW-trending fractures and related faults resulted from southeast-directed compressional stresses during the Yanshanian subduction event. During the subsequent Cenozoic extensional phase, these fractures were reactivated, creating effective storage spaces for hydrocarbons. The presence of calcite and siliceous veins within the reservoir indicates the influence of meteoric water and magmatic–hydrothermal fluid activities. Meteoric water weathering exerted a depth-dependent dissolution effect on feldspathoid minerals, leading to the formation of fracture-related pores near the top of the buried hill trap during the Mesozoic exposure period. Consequently, the combination of high-density fractures and dissolution pores forms a vertically layered reservoir within the buried hill trap. The distribution of potential hydrocarbon targets in the granitic basement is closely linked to the surrounding tectonic framework. Full article

An understanding of the textures (grain size, grain shape, porosity, etc.), composition (mineralogy), and distribution of constituent components of geological materials such as carbonate and siliciclastic sedimentary rocks is essential in their classification, interpretation, and significance in terms of their geomechanical strength and liquid/gas storage potential. In terms of scanning electron microscopy (SEM), this is limited to relatively flat areas of selected rough surfaces, or the analysis of polished thin sections. Here, we illustrate a new technique that can image large areas of the external surface of mini-cores (approximately 10 mm or smaller in diameter) drilled from carbonate and siliciclastic rock samples. The technique utilises a specially developed horizontal rotation stage within an SEM and allows the collection of high-resolution images that can be reconstructed into realistic surface representations of the mini-core surfaces. Elemental data (representative of mineralogy) can also be added using a combined backscattered electron and X-ray (BEX) detector. Currently, these reconstructions can be used as a useful tool for the analysis of both carbonate and siliciclastic geological materials. Further work may allow such reconstructions to aid in the improvement of resolution in micro-CT scans and the direct identification of mineral phases within such scans. Full article

Parallelepipeds specimens were made to further investigate the rockburst occurrence mechanism of ore pillars in underground mining units. The investigation was carried out with uniaxial compression systems and real-time testing systems, such as stress, video, and acoustic emission, combined with digital image correlation (DIC) and SEM electron microscope scanning technology, to systematically analyze the evolution of rockburst of ore pillars, strain field characteristics, acoustic emission characteristics, mesoscopic characteristics of the rockburst fracture, morphology of the bursting crater, and debris characteristics. The findings demonstrate that the pillar’s rockburst process went through four stages, including the calm period, the particle ejection period, the block spalling period, and the full collapse period. According to DIC digital image correlation technology, the development of cracks in the rock is not obvious during the calm period, but during the small particle ejection and block spalling periods, the microcracks started to form and expand more quickly and eventually reached the critical surface of the rock, resulting in the formation of a complete macro-rockburst rupture zone. During stage I of the test, the rate of acoustic emission events and energy was relatively low; from stages II to IV, the rate gradually increased; and in stage V, the rate of acoustic emission events and energy reached its maximum value at the precise moment the rock exploded, releasing all of its stored energy. The specimen pit section primarily exhibits shear damage and the fracture exhibits shear fracture morphology, while the ejecta body primarily exhibits tensile damage and the fracture exhibits tensile fracture morphology. The location of the explosion pit is distributed on the left and right sides of the middle pillar of the specimen, and the shape is a deep “V”. The majority of the rockburst debris is greater than 5 mm, and it mostly takes the shape of thin plates, which is comparable to the field rockburst debris’s shape features. Full article

The accurate identification of rock types is crucial for understanding geological structures and planning mining activities. Therefore, the precise labeling of rock types is a fundamental requirement for researchers and industry experts in these fields. This study aims to identify rock types by segmenting thin-section rock images using advanced deep learning models. Commonly used models such as DeepLabv3+, SegFormer, ConvNext, and Mask2Former were evaluated to compare the performance of different segmentation models. Additionally, an improved version of Mask2Former was analyzed to enhance its performance. Post-segmentation enhancements with SLIC super pixel and morphological operations further improved boundary delineation. An improved version of Mask2Former achieved a validation accuracy of 91.26% and a mean Intersection over Union (mIoU) of 82.59%, representing an improvement of more than 5% over the base Mask2Former. Another significant aspect of the study is the detailed analysis of other models using the relevant dataset. The Quartz Feldspar Plagioclase (QAP) diagram was utilized for mineral identification, achieving a mineral recognition accuracy of 85.7%. These results indicate the robustness of the proposed approach, which exceeds the accuracy and mIoU of comparable methods reported in the literature. This study significantly enhances the effectiveness of image processing techniques for rock type identification. Furthermore, the detailed comparison of different models provides valuable guidance for researchers in selecting the most suitable model. Full article

This study presents a novel lithological analysis method that combines optical thin-section analysis with intelligent algorithms. The method utilizes mineral composition data and two-dimensional rock properties to improve the accuracy and efficiency of lithological identification. By integrating high-resolution optical imaging to precisely capture microscopic rock textures with automated mineral composition analysis systems, this method ensures the rapid and accurate identification of major mineral components without sample damage. Furthermore, advanced image processing and similarity analysis algorithms effectively classify and automatically identify distinct rock layers, enabling a continuous and precise lithological identification process. This approach overcomes the limitations of traditional methods, including subjectivity and inefficiency, and provides robust technical support for geological exploration and mineral resource evaluation. The study shows that this method markedly improves the efficiency of petrological analysis compared to traditional logging techniques. The spatial resolution for mineral composition and lithological identification improved from 12.5 cm to 7 cm per point, maintaining an accuracy of 83%. These results underscore the method’s potential to advance technologies in geoscience and related fields. Full article

Reservoir space characteristics are the key to reservoir evaluation and the evaluation of reservoir capacity. The reservoir space of fracture-vuggy carbonate reservoirs is complex and diverse, and it develops from micro to macro. There is a lack of systematic study on the reservoir space of the Ordovician fracture-vuggy carbonate reservoir. Therefore, taking the Ordovician Yijianfang Formation in Yueman Block of Fuman Oilfield in Tarim Basin as an example, the microscopic reservoir space characteristics of the study area were characterized by rock thin section identification, X-ray diffraction, scanning electron microscopy, high-pressure mercury injection, and low-temperature nitrogen adsorption experiments, and the macroscopic reservoir space characteristics of the study area were characterized by core observation, drilling and logging data, and imaging logging data. The results showed that (1) the lithology of the Ordovician Yijianfang Formation in the Yueman area of Fuman Oilfield is mainly micrite and sparry grain limestone. The mineral composition is mainly calcite, accounting for 97.35%, containing a small amount of quartz and dolomite, accounting for 1.1% and 1.55%, respectively. (2) At the micro level, the reservoir space of Yijianfang Formation in Yueman Block is not developed in primary pores, mainly having developed dissolution pores, structural fractures, and pressure solution fractures, and the pore size is distributed from the nanometer to micron scale. (3) The dissolution caves in the study area are developed at the macro level, mainly including pore-type, cave-type, fracture-pore-type, and fracture-type reservoirs. The research results provide technical support for the accurate evaluation of fractured-vuggy carbonate reservoirs and the improvement of exploration and development effects. Full article