A method for intelligent information extraction of coal fractures based on µCT and deep learning

ObjectiveThe fine-scale characterization of fractures in coal reservoirs is significant for the exploration and exploitation of coalbed methane (CBM) resources. Given that the size, orientation, and density of fractures directly affect the permeability of coal seams, the accurate information identif...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhazha HU, Xun ZHANG, Yi JIN, Linxian GONG, Wenhui HUANG, Jianji REN, Norbert Klitzsch
Format: Article
Language:zho
Published: Editorial Office of Coal Geology & Exploration 2025-02-01
Series:Meitian dizhi yu kantan
Subjects:
Online Access:http://www.mtdzykt.com/article/doi/10.12363/issn.1001-1986.24.09.0609
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850052285091020800
author Zhazha HU
Xun ZHANG
Yi JIN
Linxian GONG
Wenhui HUANG
Jianji REN
Norbert Klitzsch
author_facet Zhazha HU
Xun ZHANG
Yi JIN
Linxian GONG
Wenhui HUANG
Jianji REN
Norbert Klitzsch
author_sort Zhazha HU
collection DOAJ
description ObjectiveThe fine-scale characterization of fractures in coal reservoirs is significant for the exploration and exploitation of coalbed methane (CBM) resources. Given that the size, orientation, and density of fractures directly affect the permeability of coal seams, the accurate information identification and extraction of fractures in coal seams plays a key role in revealing the formation and propagation mechanisms of fracture networks during reservoir volume fracturing. Conventional methods for fracture information extraction typically rely on manual labeling and feature extraction based on image processing techniques, exhibiting significantly limited accuracy and efficiency. Methods This study proposed a method for fracture information extraction of coals based on TransUNet and micro-computed tomography (µCT) images. TransUNet, integrating the advantages of both the Transformer modules and convolutional neural network (CNN), is capable of extracting global features and capturing local details in images, significantly enhancing the image segmentation accuracy and network robustness. First, the µCT images of coal samples were preprocessed, including improving the image quality using the difference method and increasing the sample size using data augmentation techniques. Subsequently, image segmentation was conducted using TransUNet to extract fracture features. Additionally, the image segmentation results of varying neural network models were compared. Results and ConclusionsThe results indicate that the proposed method exhibited superior performance on a given dataset. Specifically, the TransUNet model yielded an accuracy of 91.3%, precision of 89.5%, F1 score of 89.8%, and Intersection over Union (IoU) of 84.0%, significantly outperforming other intelligent models like U-Net and U-Net++. Given the characteristics of fine-grained µCT images, applying TransUNet to the fracture information extraction of coals emerges as an efficient and accurate approach. This study provides a novel philosophy for image processing in the field of CBM exploration and production.
format Article
id doaj-art-824c4471ec2d418f92aa8ab773db5afb
institution DOAJ
issn 1001-1986
language zho
publishDate 2025-02-01
publisher Editorial Office of Coal Geology & Exploration
record_format Article
series Meitian dizhi yu kantan
spelling doaj-art-824c4471ec2d418f92aa8ab773db5afb2025-08-20T02:52:52ZzhoEditorial Office of Coal Geology & ExplorationMeitian dizhi yu kantan1001-19862025-02-01532556610.12363/issn.1001-1986.24.09.060924-09-0609-Hu-ZhazhaA method for intelligent information extraction of coal fractures based on µCT and deep learningZhazha HU0Xun ZHANG1Yi JIN2Linxian GONG3Wenhui HUANG4Jianji REN5Norbert Klitzsch6Science and Technology R&D Platform of Emergency Management Ministry for Deep Well Ground Control and Gas Extraction Technology, School of Energy Science and Engineering, Henan Polytechnic University, Jiaozuo 454003, ChinaScience and Technology R&D Platform of Emergency Management Ministry for Deep Well Ground Control and Gas Extraction Technology, School of Energy Science and Engineering, Henan Polytechnic University, Jiaozuo 454003, ChinaSchool of Resources and Environment, Henan Polytechnic University, Jiaozuo 454003, ChinaSchool of Resources and Environment, Henan Polytechnic University, Jiaozuo 454003, ChinaSchool of Energy, China University of Geosciences (Beijing), Beijing 100083, ChinaSchool of Computer Science and Technology, Henan Polytechnic University, Jiaozuo 454003, ChinaInstitute of Applied Geophysics and Geothermal Energy, RWTH Aachen University, Aachen 52074, GermanyObjectiveThe fine-scale characterization of fractures in coal reservoirs is significant for the exploration and exploitation of coalbed methane (CBM) resources. Given that the size, orientation, and density of fractures directly affect the permeability of coal seams, the accurate information identification and extraction of fractures in coal seams plays a key role in revealing the formation and propagation mechanisms of fracture networks during reservoir volume fracturing. Conventional methods for fracture information extraction typically rely on manual labeling and feature extraction based on image processing techniques, exhibiting significantly limited accuracy and efficiency. Methods This study proposed a method for fracture information extraction of coals based on TransUNet and micro-computed tomography (µCT) images. TransUNet, integrating the advantages of both the Transformer modules and convolutional neural network (CNN), is capable of extracting global features and capturing local details in images, significantly enhancing the image segmentation accuracy and network robustness. First, the µCT images of coal samples were preprocessed, including improving the image quality using the difference method and increasing the sample size using data augmentation techniques. Subsequently, image segmentation was conducted using TransUNet to extract fracture features. Additionally, the image segmentation results of varying neural network models were compared. Results and ConclusionsThe results indicate that the proposed method exhibited superior performance on a given dataset. Specifically, the TransUNet model yielded an accuracy of 91.3%, precision of 89.5%, F1 score of 89.8%, and Intersection over Union (IoU) of 84.0%, significantly outperforming other intelligent models like U-Net and U-Net++. Given the characteristics of fine-grained µCT images, applying TransUNet to the fracture information extraction of coals emerges as an efficient and accurate approach. This study provides a novel philosophy for image processing in the field of CBM exploration and production.http://www.mtdzykt.com/article/doi/10.12363/issn.1001-1986.24.09.0609trans-unetµct imagescoal fractureimage segmentationdeep learning
spellingShingle Zhazha HU
Xun ZHANG
Yi JIN
Linxian GONG
Wenhui HUANG
Jianji REN
Norbert Klitzsch
A method for intelligent information extraction of coal fractures based on µCT and deep learning
Meitian dizhi yu kantan
trans-unet
µct images
coal fracture
image segmentation
deep learning
title A method for intelligent information extraction of coal fractures based on µCT and deep learning
title_full A method for intelligent information extraction of coal fractures based on µCT and deep learning
title_fullStr A method for intelligent information extraction of coal fractures based on µCT and deep learning
title_full_unstemmed A method for intelligent information extraction of coal fractures based on µCT and deep learning
title_short A method for intelligent information extraction of coal fractures based on µCT and deep learning
title_sort method for intelligent information extraction of coal fractures based on µct and deep learning
topic trans-unet
µct images
coal fracture
image segmentation
deep learning
url http://www.mtdzykt.com/article/doi/10.12363/issn.1001-1986.24.09.0609
work_keys_str_mv AT zhazhahu amethodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT xunzhang amethodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT yijin amethodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT linxiangong amethodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT wenhuihuang amethodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT jianjiren amethodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT norbertklitzsch amethodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT zhazhahu methodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT xunzhang methodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT yijin methodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT linxiangong methodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT wenhuihuang methodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT jianjiren methodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning
AT norbertklitzsch methodforintelligentinformationextractionofcoalfracturesbasedonμctanddeeplearning