PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
April 14, 2026Bulletin of Mineralogy Petrology and Geochemistry0 citations

A Geological Logic-Informed AI Method for Oolitic Limestone Identification

View Full Paper
CSCheng SuYLYifan LiaoYLYijie Liu

Key Points

  • This work aims to develop an AI method for the accurate identification of oolitic limestone using geological knowledge.
  • Developed an AI-based method for oolitic limestone identification.
  • Divided the identification process into four key stages: particle recognition, pore pixel extraction, filler classification, and rock naming.
  • Utilized deep learning to improve identification accuracy and reliability.
  • Incorporated geological logic to make the AI's process transparent and interpretable.
  • Achieved an identification accuracy rate of 91.8% on the test set.
  • Provided interpretable and traceable results for each stage of the identification process.
  • Established a logical transparency and reviewable knowledge basis for rock naming.

Abstract

鲕粒灰岩(oolitic limestone)是一种典型的浅水高能环境碳酸盐沉积岩,其微观属性是反演古环境动力条件、解析成岩演化历程及评价储层物性的核心指标,因此其显微镜下的分析鉴定是油气勘探及相关研究的重要基础工作。传统岩石薄片人工鉴定方法存在耗时长、主观性强及重现性差等瓶颈,难以适应海量地质数据的分析需求。为此,本文提出一种基于地质逻辑驱动的鲕粒灰岩人工智能自动鉴定方法,基于专家鉴定知识将鲕粒灰岩的鉴定过程划分为鲕粒颗粒识别、孔隙像素级提取、填隙物自动分类,以及基于规则集的岩石定名4个主要环节,将深度学习的“黑盒”预测转化为受地质先验知识约束的可解释鉴定流程,有效提升了结果的准确性与可靠性。实验结果表明,该方法在测试集上的鉴定准确率为91.8%,并且可以得到鉴定过程中各环节的可解释、可追溯的分项结果及鉴定依据,使岩石定名过程具备逻辑透明性和知识可审查性,从而为岩石薄片图像的智能化分析提供了兼具可解释性与工程应用潜力的创新思路。

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Su et al. (2026) studied this question.

synapsesocial.com/papers/69ddd8eee195c95cdefd6795https://doi.org/10.3724/j.issn.1007-2802.20260081
Ask AI
Helpful
Bookmark
Share
View Full Paper