69 results about "Thin section" patented technology

Systems and methods for determining a diagenesis level in a reservoir for performing hydrocarbon extraction include receiving a set of open hole (OH) log data from one or more wells in the reservoir, the OH log data representing a subsurface of the reservoir; executing a diagenesis model to process the set of OH log data, the diagenesis model trained by thin-section image data correlated to labeled OH log data, the labeling identifying values for a quartz overgrowth rate (QOR), a clay coating rate (CCR), or a thin-section porosity in the subsurface based on a value in the OH log data; determining a prediction of a porosity of the reservoir, a permeability of the reservoir, or both the porosity and the permeability; and generating a control signal representing a recommendation to drill a well at a location in the reservoir.

The invention relates to the technical field of material analysis, in particular to a light microscope-field emission electron microscope combined shale rock slice analysis and identification method, which comprises the following steps of: firstly, combining a brightness distribution condition represented by a bright field image of a shale rock slice and a birefringence deviation condition represented by a cross polarization image from an optical angle; determining a mineral discrimination index representing an atomic number agent; applying offsets of different sizes based on the distribution condition of mineral discrimination indexes and the surface fluctuation condition of shale rock slices, and determining a position offset vector by combining gray scale gradient distribution on the basis; and finally, determining the cross-scale positioning and more accurate electron microscope objective table coordinates corresponding to each pixel point in combination with a position offset vector in a pixel coordinate conversion process based on an image physical distance scale and rotation translation processing, so that the cross-scale positioning accuracy of the optical microscope and the field emission scanning electron microscope is higher. The reliability of analyzing and identifying the shale rock slice is obviously improved.

The invention discloses a rock slice image component identification method based on deep learning. According to the method, a data set is constructed through PCA dimension reduction and normalization, segmentation parameters are generated by utilizing ResNet50 and a self-attention mechanism, automatic classification of quartz, rock debris, kaolinite, pores and matrixes is realized by combining a VGG16 model, two-stage segmentation is performed by adopting AGLFIF and Otsu algorithms, and high-precision identification and geological analysis automation of rock slice images are realized. According to the method, real-time component identification of the rock slice image can be realized, and automation and reliability of geological analysis are improved.

The invention discloses a method for in-situ leaching mining of sandstone uranium ore, the concentration of bicarbonate radical ions in ore bed water in the sandstone uranium ore is lower than 0.5 g / L, the sandstone uranium ore comprises yttrium-containing uranium ore, and the carbonate content is 1.5%-2.5%; the ore bed water serves as a solvent, 500 mL of leaching agents with different pH values adjusted through sulfuric acid are prepared, and the concentration of hydrogen peroxide in the leaching agents is 2 g / L; and determining the maximum pH value when the uranium ore in the slice sample is corroded by using the leaching agents with different pH values, and taking the maximum pH value as a critical pH value. The leaching agent with the critical pH value is used for exploiting the sandstone uranium ore, corrosion to gangue minerals in the leaching process can be reduced, the number of formed colloids and CaSO4 precipitates is reduced, the precipitates are prevented from blocking an ore bed, in addition, the method can be used for target corrosion of the refractory uranium ore, and therefore efficient leaching is achieved.

The invention discloses a method and a device for quantitatively characterizing the influence intensity of ancient volcanic activities, which can be used in the field of geophysical exploration, and the method comprises the following steps: determining the compaction recovery coefficient of vacant tuff according to the sorting coefficient and the current porosity; according to the compaction recovery coefficient and the thickness of the bare slice hollow falling type tuff, the original deposition thickness of the hollow falling type tuff is determined; according to the original deposition thickness and the number of layers of the empty tuff, determining the volcanic activity influence intensity of the to-be-detected area; wherein the sorting coefficient is the sorting coefficient of the empty tuff and is obtained by carrying out particle size analysis on a polished thin section; the current porosity is the current porosity of the empty tuff corresponding to the polished section and is obtained by measuring the porosity of the polished section; the thickness and the number of layers of the empty tuff corresponding to the bare slice are obtained by identifying the bare slice. According to the method, evaluation standards of the volcanic activity influence can be established, and the intensity of the volcanic activity influence on the to-be-detected area can be quantitatively represented.

The invention discloses a volcanic rock lithology identification method based on latent attribute characterization and Bayesian inference. According to the method, volcanic rock core / rock debris slice identification and related geochemical data are taken as calibration basis, multiple groups of logging curves sensitive to the lithologic characteristics of volcanic rock are preferably selected, and quality control and normalization processing are carried out; the method comprises the following steps: learning from a multi-logging curve through an encoder-decoder to obtain a group of mutually decoupled latent attribute representations, and respectively constructing a silicon dioxide content Bayesian classification model and a volcanic rock cause Bayesian classification model by taking the latent attributes as input; the probability prediction based on the silicon dioxide content range (basic / neutral / acidic) and the volcanic rock cause type is realized, the two prediction results are combined, and finally the volcanic rock lithology of the corresponding well section is determined. Meanwhile, the quantitative output uncertainty is used for multi-solution risk prompt, so that the accuracy, the stability and the credibility of a prediction result are improved.

The invention belongs to the field of basic research of sandstone type uranium ore associated resource evaluation, and particularly relates to a method for judging the existence form of associated gallium in a sandstone type uranium deposit, which comprises the following steps of: measuring the contents of major elements and trace elements of a plurality of uranium ore samples, and judging whether gallium meets the lowest grade requirement of associated resources of the uranium deposit; for uranium ore samples exceeding the lowest grade requirement of gallium associated resources, correlation analysis of gallium elements, main elements and trace elements is carried out, and possible key mineral types closely related to gallium are preliminarily speculated; carrying out slice analysis, and circling an approximate range of speculated possible key mineral types; carrying out scanning electron microscope and energy spectrum analysis combined observation, and determining the existence form of gallium in the uranium deposit; the position exceeding the associated grade gallium is determined through laser in-situ analysis; the gallium-related mineral type is directly determined through electron probe analysis; and comprehensively judging the existence form of gallium in the uranium deposit. According to the method, the existence form of associated gallium in the sandstone type uranium deposit can be accurately and quickly judged.

The application provides a rock slice image data management method, device and equipment, and relates to the technical field of geological exploration. The method comprises the following steps: acquiring image data and original identification data of a rock slice to be managed, which have first annotation information; identifying key features of the rock through machine learning based on the first annotation information and an identification standard to obtain second annotation information, wherein the second annotation information contains information carrying a first annotation information label; preliminarily checking the original identification data by using the second annotation information carrying the first annotation information label to obtain a checking result and annotation deviation information; performing secondary checking on the original identification data by using the second annotation information not carrying the first annotation information label to obtain annotation missing information; and managing the image data based on the checking result, the annotation deviation information and the annotation missing information in combination with double-expert collaborative identification. The application can improve the accuracy of rock slice image data management.

This invention relates to the field of carbonate rock lithology identification technology, and particularly to a method, system, and apparatus for automatically identifying carbonate rock lithology based on well logging curves; comprising the following steps: S1, acquiring well logging curve data, core data, and thin section data from multiple wells in the area to be identified; S2, standardizing and environmentally correcting the well logging curve data; S3, using an optimization model to solve for the lithological profile of the area to be identified, thereby determining the clay content, porosity, and water saturation; S4, constructing rock structure curves and formation fracture porosity curves; S5, constructing a machine learning model, setting a learning sample set, and inputting the learning sample set into the machine learning model for training, so that the trained machine learning model can automatically identify carbonate rock lithology; this invention improves the accuracy of carbonate rock lithology identification.

The embodiment of the application provides a preparation method of a fluid inclusion microthermometric ore slice, comprising the following steps: providing a fluid inclusion slice which has been observed and circled under an optical microscope in detail for microthermometric target fluid inclusion position; putting the whole fluid inclusion slice into a glass culture dish, injecting a slice adhesive dissolving solution into the culture dish, and immersing the whole fluid inclusion slice in the slice adhesive dissolving solution to dissolve the slice adhesive and separate the fluid inclusion ore slice from the bonding surface of the glass slide; after the separation, placing the fluid inclusion ore slice with the adhesive surface upward on the surface of the glass slide and immersing it in the slice adhesive dissolving solution; after the slice adhesive on the surface of the fluid inclusion ore slice is completely dissolved, cutting the fluid inclusion ore slice according to the position of the target inclusion to be measured and thermometric, and preparing a rock and mineral fragment containing the target fluid inclusion which is suitable in size to be placed in a cold and hot stage quartz crucible.

The invention relates to the technical field of strata facies type identification, in particular to a method and device for rapidly identifying shale strata facies, and the method comprises the following steps: obtaining a to-be-detected strata-containing shale sample; the method comprises the following steps: scanning a to-be-detected shale sample containing the strata by using a portable handheld X-ray fluorescence spectrometer, constructing a measuring point position, and measuring elements in the sample point by point at the measuring point position to obtain the mass percent content of the elements; performing contrast microscopic analysis, selecting corresponding part of point positions at the element measuring point positions to prepare rock slices, and identifying the exact lamina phase type of the rock slices under a microscope; according to the method, the element content in the to-be-detected shale sample containing the lamina is measured through the portable handheld X-ray fluorescence spectrophotometer, the lamina phase type is judged and identified through the element content, identification of a single lamina can be completed within one minute, identification of the lamina phase type can be rapidly completed in a field environment in a lossless and low-cost mode, and the identification accuracy is high. And the identification efficiency of the lamina phase type is improved.

The invention provides a carbonate rock seismic lithology depicting method and system, a storage medium and equipment. The method comprises the following steps: collecting basic geological data of a research area; lithology identification is carried out on the rock core and the slice of the target layer based on drilling and logging data, the calibration relation between different lithology and a logging curve is obtained, and the content of different lithology in the target layer is calculated; establishing a through-well seismic synthesis record based on the logging curve data, performing target stratum top and bottom stratum tracking in the three-dimensional seismic data volume, and extracting the internal seismic attributes of the target stratum; performing correlation analysis on the contents of the different lithologies and the seismic attributes, obtaining a correlation coefficient between the lithology contents and the seismic attributes, selecting the seismic attributes based on the correlation coefficient, and determining a lithology regression equation; and calculating and acquiring content ratio results of different lithologic components, and integrating the content ratio results to obtain stratum lithofacies distribution characteristics of the target stratum. According to the carbonate rock seismic lithology depicting method, the depicting accuracy is improved.

The invention discloses a rock slice image classification method based on domain self-adaption, and the method comprises the steps: collecting rock slice image data, the method comprises the following steps: acquiring a rock slice image, performing multi-scale data preprocessing and geological field data enhancement, extracting multi-level visual features of the rock slice image by utilizing a pre-trained DINOv3 model, performing field specialized adaptation on general visual features through a geological field adaptive Adapter module, and enhancing rock slice discriminative feature representation by adopting a double-path attention mechanism. Constructing a progressive hierarchical classification head to realize coarse-to-fine rock classification; and designing a multi-stage progressive training strategy to optimize the overall performance of the model. According to the method, mineral composition and structural features of the rock slices under different scales can be accurately captured, and multi-scale features and an attention mechanism are fully utilized, so that accurate classification of the rock slices is realized, and the accuracy and reliability of rock slice identification are remarkably improved.

This invention relates to the field of petroleum exploration and development technology, and is a method and system for fluid property identification based on micro-fluorescence thin section technology. The method includes acquiring the micro-fluorescence thin section spectrum of a rock sample; obtaining the true fluorescence area under blue light and the true fluorescence area under ultraviolet light based on the micro-fluorescence thin section spectrum of the rock sample; substituting the true fluorescence area under blue light and the true fluorescence area under ultraviolet light into an extinction index model to obtain an extinction index; and comparing the extinction index with an extinction index evaluation standard to identify fluid properties. This invention establishes an extinction index for identifying reservoir fluid properties. Its extinction index model extracts the pair of sensitive parameters—fluorescence area under blue light and ultraviolet light—by utilizing the spectral variation characteristics between different crude oils and different proportions of water under micro-fluorescence thin section technology. This effectively solves the problem of reservoir fluid property identification, and the method can quantitatively identify reservoir fluid properties.

Systems and methods for determining a diagenesis level in a reservoir for performing hydrocarbon extraction include receiving a set of open hole (OH) log data from one or more wells in the reservoir, the OH log data representing a subsurface of the reservoir; executing a diagenesis model to process the set of OH log data, the diagenesis model trained by thin-section image data correlated to labeled OH log data, the labeling identifying values for a quartz overgrowth rate (QOR), a clay coating rate (CCR), or a thin-section porosity in the subsurface based on a value in the OH log data; determining a prediction of a porosity of the reservoir, a permeability of the reservoir, or both the porosity and the permeability; and generating a control signal representing a recommendation to drill a well at a location in the reservoir.

A method and a system for generating a labeled benchmark dataset are disclosed. The method includes obtaining a plurality of sources related to a thin section using web scraping and extracting a plurality of images from the plurality of sources related to the thin section, the plurality of images including a plurality of thin section images and a plurality of non-thin section images. Further, the method includes determining the plurality of thin section images from the plurality of extracted images and generating a classification of the plurality of thin section images based on a given classification criteria. The geological thin-section based machine learning models is trained based on the generated classification of the plurality of thin section images and a wellbore drilling plan is generated based on the geological thin-section based machine learning models.

This invention relates to the field of materials analysis technology, specifically to a method for analyzing and identifying shale thin sections using a combination of optical microscopy and field emission scanning electron microscopy. First, from an optical perspective, the method combines the brightness distribution of bright-field images of shale thin sections with the birefringence deviation of cross-polarized images to determine mineral discrimination indices representing atomic number proxies. Then, based on the distribution of the mineral discrimination indices and the surface undulations of the shale thin sections, different biases are applied, and the positional bias vector is determined by combining this with the gray-level gradient distribution. Finally, based on the physical distance scale of the image and the pixel coordinate transformation process involving rotation and translation, the positional bias vector is used to determine the electron microscope stage coordinates corresponding to each pixel point for more accurate cross-scale positioning. This improves the accuracy of cross-scale positioning using optical microscopy and field emission scanning electron microscopy, significantly enhancing the reliability of shale thin section analysis and identification.

The invention provides an intelligent sandstone slice identification system and an intelligent sandstone slice identification method, and belongs to the technical field of geological exploration. In the system, a data analysis module integrates three core algorithms: a data enhancement module adopts an improved conditional cycle consistency generative adversarial network to solve the problem of scarcity of rock samples; the multi-modal feature understanding module fuses image visual features and petrology text features by using a cross-modal attention mechanism; and the intelligent identification module completes recognition and classification of mineral composition, structure construction and diagenesis in parallel based on a multi-task learning architecture. The method comprises a system training stage and a sample identification stage. According to the method, the model generalization ability is improved by generating a large number of synthetic samples, the geological rationality of the identification result is enhanced through multi-modal semantic fusion, efficient, accurate and objective automatic intelligent identification of the sandstone slice is realized, and a key technical support is provided for reservoir evaluation in the field of petroleum geology.

The application discloses a kind of carbonate rock flake identification method and system, belong to image processing technical field, will take the carbonate rock flake image of different geological scene respectively input pre-trained network model and identify, output attribute identification result.Network model includes classification module, segmentation module, feature fusion module and classifier, after classification module extracts mineral texture feature of carbonate rock flake image by mineral texture attention module, respectively input classification head and the regression head of embedding dynamic scaling constraint regression mechanism and carry out detection;Segmentation module carries out regional segmentation to carbonate rock flake image, outputs the pixel proportion of real mask corresponding to the prediction mask of different segmentation region respectively;Feature fusion module carries out fusion to detection result and pixel proportion;Classifier carries out coordinated inference to fusion feature, and outputs attribute identification result.The method is identified by cooperation and dynamic constraint, ensure the scientificity and reliability of identification result.

The invention discloses a high-precision recovery marine facies compact limestone sedimentary environment and oil-gas dessert area prediction method. The method comprises the steps that S1, an argillaceous limestone phase, an argillaceous limestone phase, a mud-powder crystal limestone phase and a paste limestone phase are divided through X-ray diffraction and slice identification; s2, utilizing a chemical weathering index CIA, a Sr / Ca salinity index and delta 13Corg organic carbon isotope to synergistically recover paleo-climate and paleo-hydrological conditions; s3, combining total organic carbon TOC, organic matter maturity Ro, terrestrial input Al and Ti content and nanopore development degree, and establishing an organic matter-rich high-quality lithofacies evaluation standard; and S4, according to an evaluation standard, well drilling, well logging and seismic data are fused to predict the oil and gas sweet spot area. Through the four steps of lithofacies division, paleoenvironment recovery, evaluation standard establishment and dessert area prediction, high-precision recovery of the marine facies compact limestone sedimentary environment is realized, the dessert area prediction success rate is improved to 78%, and efficient technical support is provided for oil-gas exploration.

The invention discloses an ore structure feature recognition method based on deep learning, particularly relates to the technical field of intelligent recognition of ore structure features, and is used for solving the problem of how to construct a multi-modal deep learning recognition method capable of running on an edge side under the condition of lacking large-scale same-domain labeling and unified sample preparation. And the stable identification of crossing mining areas, crossing equipment and crossing sample preparation is realized. A slice, a hand specimen, a rock core image, hyperspectral data and an SEM-EDS element graph are collected at the same time in a mining area, coarse registration, fine registration and scale alignment are completed under a unified coordinate system, and self-supervised training, domain self-adaption and multi-modal segmentation are combined to obtain a high-resolution image. According to the method, the ore particles, the contact relation and the crack structure can be stably recognized under different mining areas, different sample preparation conditions and different collection devices, and a redetectable structured quantitative index set is output, so that the problems of unstable cross-scene recognition and non-traceable result in the prior art are solved.

The present application relates to a rock porosity calculation method and system based on clastic rock thin section identification data, and belongs to the technical field of oil and gas exploration and geology engineering. The method comprises the following steps: obtaining a first compaction rate according to the carbonate content data in the current clastic rock thin section identification data and the fitting relationship between the carbonate content and the compaction rate in the historical data; obtaining a second compaction rate according to the clay contribution rate obtained from the current clastic rock thin section identification data and the fitting relationship between the clay contribution rate and the compaction rate obtained from the historical data; obtaining a comprehensive compaction rate by weighted average of the first and second compaction rates; obtaining a modified compaction rate by modifying the comprehensive compaction rate according to the ratio of clay to initial porosity and the ratio of carbonate to initial porosity; determining the calculated porosity by using the modified compaction rate; and obtaining the rock porosity by modifying the calculated porosity according to the dissolution and cementation of the thin section. The clay contribution rate is the ratio of the square root of the clay content to the sum of the clay content and the carbonate content.

The invention discloses a reservoir high-quality lithology analysis method and system for metamorphic rock type bedrock, and relates to the technical field of oil and gas geological reservoir analysis, and the method is characterized in that a rock slice image of the metamorphic rock type bedrock collected under a single polarizing microscope is acquired; based on the rock slice image, identifying and classifying different lithologic reservoirs, and extracting secondary reservoir space features in each lithologic reservoir; based on the reservoir space characteristics, determining the total surface porosity and the effective reservoir space proportion of each lithologic reservoir; and comparing and judging the total surface porosity and the effective reservoir space ratio of each lithologic reservoir through the lithologic identification model to obtain an identification result of the high-quality lithologic characters of the reservoirs. According to the method, whether the lithology of various metamorphic bed rock reservoirs is high or not can be effectively and accurately judged, the reservoir performance of the metamorphic bed rock reservoirs with different lithology is converted into visual objective data, and the production and operation cost is effectively reduced.

The invention discloses a rock slice image segmentation method and device based on machine learning and a medium, and relates to the technical field of geological exploration image processing, a rock slice image is divided into a plurality of initial sub-regions, textures in each initial sub-region are analyzed, a texture primitive shape sequence is constructed, and according to the texture primitive shape sequence, a rock slice image is segmented. According to the method, data support can be provided for subsequent analysis of texture similarity between any two initial sub-regions, and in combination with the texture primitive distribution difference between any two initial sub-regions, the proximity degree of the two initial sub-regions on texture features can be reflected, and the segmentation boundary and region division precision of the image can be further improved.

This application provides an experimental method for testing the permeation efficiency of mixed sedimentary rock cores, including: cutting the mixed sedimentary rock core into thin sections for clay mineral analysis; weighing a standard core and recording the weight as m1; performing a displacement experiment on the standard core and saturating it with oil; immersing the standard core in oil and aging it in a constant temperature chamber, then weighing it and recording the weight as m2; measuring the T2 spectrum of the standard core at its spatial location and calculating the peak area S1; weighing the heat-shrink tubing and recording the weight as m0; wrapping the standard core circumferentially with heat-shrink tubing, placing it in a hanging ring and suspending it on an electronic balance, immersing it in the permeation fluid, and recording the weight data at permeation equilibrium as m3; measuring the T2 spectrum of the standard core at its spatial location and calculating the peak area S2; calculating the permeation efficiency based on the gravimetric method; calculating the permeation efficiency using the nuclear magnetic resonance peak areas before and after the permeation experiment; and calculating the permeation efficiency of the mixed sedimentary rock core. This method can accurately guide the effective development of mixed sedimentary rock reservoirs.

This invention relates to a method for calculating the critical pressure gradient of water-sealed gas based on scanning electron microscope (SEM) image data, comprising: step S1, acquiring thin-section images of rocks in water-sealed gas formations using an SEM; step S2, extracting the thin-section image data and binarizing the images to obtain a binarized image of the thin-section; step S3, converting the binarized image into a vectorized pore-throat graph, analyzing the vectorized pore-throat graph, characterizing the microstructure of the pore-throat, and obtaining pore-throat structural parameters; step S4, establishing the relationship between pore-throat structural parameters and flow parameters using boundary layer theory to obtain the critical pressure gradient of water-sealed gas. This invention, based on SEM images of rocks, uses boundary layer theory to obtain the critical pressure gradient of water-sealed gas, quantitatively evaluating the critical pressure gradient parameters for gas to overcome water lock, and providing a theory and method for releasing the water-sealed gas effect in gas reservoirs.

The embodiment of the invention discloses an ore indoor in-situ solution corrosion and evaluation method. The method comprises the steps that blocky ore is made into a cube or cylinder or sheet-shaped sample; one surface of the blocky ore is a natural split surface, and the other surface of the blocky ore is a polished surface; carrying out mineralogy and geochemical characteristic research on the prepared sample before leaching; determining reagent types and concentration conditions of the leaching solution; determining the leaching time interval and the leaching times; leaching the ore sample subjected to the research before leaching, and stirring through an automatic stirrer in the leaching process; the ore sample reaching the leaching time is taken out, and mineralogy and geochemical characteristic research is carried out till the leaching times are completed; comparing mineralogical and geochemical characteristics of the ore before leaching and after each leaching time; according to the experimental method, the petrology characteristics of the ore can be represented, and the change rules of mineralogy, geochemical characteristics and leaching liquid chemical composition of the ore at different leaching stages can be compared.

The invention relates to the technical field of geological evaluation, in particular to a stratified pulveryte core geology rapid experimental evaluation method which comprises the following steps: establishing a sample lithology characteristic table according to the color, mineral type and sedimentary structure of a stratified pulveryte core; determining a sampling layer section of the lithologic layer section of the lamina type pulveryte rock to be evaluated, and collecting and preparing a rock core sample; testing the collected rock core sample to obtain test data, wherein the test comprises slice identification, total rock X diffraction analysis, porosity determination and pyrolysis analysis experiment; the test data is extracted, homing comparison is carried out according to the test data and the sampling position, and core homing processing of the test data is achieved; and evaluating and analyzing the core potential of the lamina type pulveryte rock based on test data. According to the method, the rapid evaluation of the hydrocarbon source rock and the oiliness is realized, the sampling heterogeneity problem in the evaluation process is solved, and guidance is provided for the fine evaluation of the shale oil reservoir.

The invention relates to the technical field of geological exploration, in particular to a method for obtaining the age of Rb-Sr in a potassium-rich mineral microcell. The technical operation process is simple, and complicated chemical pretreatment is not needed; various sample carriers (such as sample targets or rock slices and the like) can be directly analyzed, and high spatial resolution (50-100 microns) is achieved; the analysis speed is high (1t, 2 minutes); the Rb-Sr dating result has relatively high accuracy (100.1%-101.6%) and precision (superior to + / -2.3%).

The invention relates to the technical field of geological lithology intelligent identification, and discloses a limestone microscopic image intelligent identification method, which comprises the following steps: extracting a single polarized light image and an orthogonal polarized light image of a rock slice, and carrying out lithology marking on the single polarized light image and the orthogonal polarized light image to obtain a rock slice image data set; a Faster Net model is trained through a rock slice image data set, feature extraction is carried out on the rock slice image data set through an ScConv-PConv-PConv-ScConv convolution structure in the optimized Faster Net model, a DualSE attention mechanism is inserted into the optimized Faster Net model, and feature fusion is carried out on extracted features through the DualSE attention mechanism; and the rock image to be identified is input into the optimized FasterNet model, so that the accuracy can be quickly improved in short-time training, the convergence process of the model is accelerated, and the lithology identification accuracy is effectively improved.