A Segnet-Based Improved Algorithm for Pore Extraction in Rock Thin Section Images

WU Yi-fei; TENG Qi-zhi; HE Hai-bo

doi:10.12263/newIT.2025.02.005

您当前的位置：

首页 >

文章列表页 >

A Segnet-Based Improved Algorithm for Pore Extraction in Rock Thin Section Images

RESEARCH PAPER | 更新时间：2025-12-01

- A Segnet-Based Improved Algorithm for Pore Extraction in Rock Thin Section Images
- New Generation of Information Technology Vol. 8, Issue 2, Pages: 24-32(2025)
- 作者机构：
  
  1.四川大学电子信息学院，四川成都 610065
  2.成都西图科技有限公司，四川成都 610065
- 作者简介：
- 基金信息：
- DOI：10.12263/newIT.2025.02.005
  CLC： TP391.41
- Revised：2025-03-24，
  
  Accepted：31 March 2025，
  
  Published：15 February 2025
- 稿件说明：
移动端阅览
吴一飞, 滕奇志, 何海波. 基于改进Segnet的岩石薄片图像孔隙提取算法[J]. 新一代信息技术, 2025, 8(2): 24-32

WU Yi-fei, TENG Qi-zhi, HE Hai-bo. A Segnet-Based Improved Algorithm for Pore Extraction in Rock Thin Section Images[J]. New Generation of Information Technology, 2025, 8(2): 24-32
吴一飞, 滕奇志, 何海波. 基于改进Segnet的岩石薄片图像孔隙提取算法[J]. 新一代信息技术, 2025, 8(2): 24-32 DOI： 10.12263/newIT.2025.02.005.

WU Yi-fei, TENG Qi-zhi, HE Hai-bo. A Segnet-Based Improved Algorithm for Pore Extraction in Rock Thin Section Images[J]. New Generation of Information Technology, 2025, 8(2): 24-32 DOI： 10.12263/newIT.2025.02.005.

摘要

岩石铸体薄片的孔隙特征分析在石油地质部门中被广泛应用于评估岩石储层质量。传统的分割岩石铸体薄片图像孔隙的方法，如色彩阈值分割法，在分割岩石铸体薄片图像孔隙时存在精度较差、需要人工大量调整的问题。而目前用于孔隙提取的深度学习神经网络在面对低孔隙度、孔隙边缘不太清晰的铸体图像时效果不太理想。针对这些问题，本文在Segnet网络的基础上，提出了一种优化后的注意力机制并结合CBAM（Convolutional Block Attention Module）模块的方法。该方法尽可能保留特征图中的空间细节，并使网络能够在空间维度上建立远距离依赖关系。本文采用目前实际生产中常见的多种铸体岩石薄片进行对比实验，选取像素精准度、平均交并比以及处理时间这三项指标来进行对比分析，结果显示本文方法均取得较好指标，表明本文方法在铸体提取时有着更好的精确度。

Abstract

The analysis of pore characteristics in rock cast thin sections has been widely applied in petroleum geology for evaluating reservoir quality. Current traditional methods

such as color threshold segmentation

exhibit limitations including low accuracy and significant manual intervention requirements when segmenting pores in cast thin section images. Meanwhile

existing deep learning neural networks for pore extraction demonstrate suboptimal performance when processing low-porosity cast images with blurred pore boundaries. To address these challenges

this study proposes an enhanced network architecture based on Segnet

incorporating an optimized attention mechanism integrated with the convolutional block attention module (CBAM).The proposed method effectively preserves spatial details in feature maps while establishing long-range dependencies across spatial dimensions.Comparative experiments are conducted using diverse cast rock thin sections commonly employed in industrial practice. Evaluation metrics including pixel accuracy (PA)

mean intersection over union (mIoU)

and processing time demonstrate the superiority of the proposed method over conventional approaches

confirming its enhanced precision in pore extraction tasks.

关键词

Keywords

references

BING H G , RONG G X , CAI Y B . Automatic thresholding of gray-level images using two-dimensional Bayes discrimination rule [J ] . Proceedings of SPIE - The International Society for Optical Engineering , 2001 , 252 - 257 .

宋震 . 基于数字图像处理技术的岩石孔隙参数提取方法研究 [D ] . 南京 : 南京大学 , 2014 .

MOLLAJAN A , GHIASI-FREEZ J , MEMARIAN H . Improving pore type identification from thin section images using an integrated fuzzy fusion of multiple classifiers [J ] . Journal of Natural Gas Science and Engineering , 2016 , 31 : 396 - 404 .

代贺 , 滕奇志 , 伦增珉 . 岩石铸体薄片图像孔隙自动提取方法 [J ] . 图像与信号处理 , 2017 , 6 ( 1 ): 17 - 28 .

李周 . 岩石铸体薄片图像的孔隙提取及颗粒分割研究 [D ] . 成都 : 四川大学 , 2018 .

蔡宇恒 , 滕奇志 , 涂秉宇 . 基于深度学习的岩石铸体薄片图像孔隙自动提取 [J ] . 科学技术与工程 , 2020 , 20 ( 28 ): 11685 - 11692 .

CAI Y H , TENG Q Z , TU B Y . Automatic extraction of pores in thin slice images of rock castings based on deep learning [J ] . Science Technology and Engineering , 2020 , 20 ( 28 ): 11685 - 11692 . (in Chinese)

刘凯文 , 熊淑华 , 滕奇志 . 基于U-net的铸体岩石薄片图像孔隙自动提取 [J ] . 智能计算机与应用 , 2021 , 11 ( 11 ): 68 - 75 .

LIU K W , XIONG S H , TENG Q Z . Automatic extraction of pores in slice image of rock castings based on U-net [J ] . Intelligent Computer and Applications , 2021 , 11 ( 11 ): 68 - 75 . (in Chinese)

BADRINARAYANAN V , KENDALL A , CIPOLLA R . SegNet: a deep convolutional encoder-decoder architecture for image segmentation [J ] . IEEE Transactions on Pattern Analysis and Machine Intelligence , 2017 , 39 ( 12 ): 2481 - 2495 .

王翔 , 史操 , 袁正一 . 融合分区注意力UNet模型用于分割MRI中的膝关节软骨 [J ] . 中国医学影像技术 , 2024 , 40 ( 5 ): 764 - 768 .

WANG X , SHI C , YUAN Z Y . UNet fusing patch attention for segmenting knee cartilage on MRI [J ] . Chinese Journal of Medical Imaging Technology , 2024 , 40 ( 5 ): 764 - 768 . (in Chinese)

PANG B . Classification of images using EfficientNet CNN model with convolutional block attention module (CBAM) and spatial group-wise enhance module (SGE) [C ] // International Conference on Image, Signal Processing, and Pattern Recognition (ISPP 2022) . SPIE , 2022 : 12247 .

何自芬 , 史本杰 , 张印辉 , 等 . 多注意力融合的环高原湖泊遥感影像分割 [J ] . 电子学报 , 2023 , 51 ( 4 ): 885 - 895 .

HE Z F , SHI B J , ZHANG Y H , et al . Remote sensing image segmentation of around plateau lakes based on multi-attention fusion [J ] . Acta Electronica Sinica , 2023 , 51 ( 4 ): 885 - 895 . (in Chinese)

ZHANG Y , LIU C , LIU M , et al . Attention is all you need: Utilizing attention in AI-enabled drug discovery [J ] . Brief Bioinform , 2023 , 25 ( 1 ): bbad467 .

杨宇 . 融合Transformer架构基于改进注意力机制的目标检测技术研究 [D ] . 北京 : 北京邮电大学 , 2024 .

俞奋孝 . 注意力机制在提高YOLOv5目标检测精度的研究 [D ] . 阜阳 : 阜阳师范大学 , 2023 .

郑云飞 , 王晓兵 , 张雄伟 , 等 . 基于金字塔知识的自蒸馏HRNet目标分割方法 [J ] . 电子学报 , 2023 , 51 ( 3 ): 746 - 756 .

ZHENG Y F , WANG X B , ZHANG X W , et al . The self-distillation HRNet object segmentation based on the pyramid knowledge [J ] . Acta Electronica Sinica , 2023 , 51 ( 3 ): 746 - 756 . (in Chinese)

SUDRE C H , LI W Q , VERCAUTEREN T , et al . Generalised dice overlap as a deep learning loss function for highly unbalanced segmentations [C ] // Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support . Cham : Springer International Publishing , 2017 : 240 - 248 .

岳枫云 , 何小海 , 滕奇志 , 等 . 基于改进UNet++的铸体薄片图像喉道分割算法 [J ] . 新一代信息技术 , 2023 ( 23 ): 15 - 22 .

YUE F Y , HE X H , TENG Q Z , et al . An improved UNet++ based algorithm for throat segmentation in cast slice images [J ] . New Generation of Information Technology , 2023 ( 23 ): 15 - 22 . (in Chinese)

LI Z , ZHANG H , LI Z Z , et al . Residual-attention UNet++: A nested residual-attention U-Net for medical image segmenta- tion [J ] . Applied Sciences , 2022 , 12 ( 14 ): 7149 .

高凤 , 吴艺妮 , 褚诗伟 . 基于FCN的计算机图像区域定位算法研究 [J ] . 鄂州大学学报 , 2024 , 31 ( 3 ): 96 - 98 .

GAO F , WU Y N , CHU S W . A research on computer image region localization algorithm based on FCN [J ] . Journal of Ezhou University , 2024 , 31 ( 3 ): 96 - 98 . (in Chinese)

汤红忠 , 王蔚 , 王涛 , 等 . 一种基于课程学习的胚胎图像语义分割方法 [J ] . 电子学报 , 2023 , 51 ( 11 ): 3365 - 3376 .

TANG H Z , WANG W , WANG T , et al . A semantic segmentation method of embryo image based on curriculum learning [J ] . Acta Electronica Sinica , 2023 , 51 ( 11 ): 3365 - 3376 . (in Chinese)

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Bone Age Prediction Based on Lightweight VGG16 and Attention Mechanism

Multi-Modality Cross Gait Recognition Combined with Transformer

Related Author

GUO Wei

GUO Jin-xing

CHEN Guang-xin

HAN Xue-shan

QIN Yu-zhen

LIU Yi-xin

QING Lin-bo

HE Xiao-hai

Related Institution

College of Life Science， Mudanjiang Medical University

Hongqi Hospital Affliated to Mudanjiang Medical University

School of Medical Imaging， Mudanjiang Medical University

National Clinical Research Center for Geriatrics, West China hospital, Sichuan University

⁰