Scholarlink

Abstract <p>A method is developed for automated identification of groups of macerals (vitrinite, inertinite, liptinite, and semivitrinite) and mineral inclusions from digital microphotographs of polished coal briquet surfaces. It is based on convolutional neural networks (CNN) and computer vision. For a test sample, the accuracy attained is 92.31%; that is comparable with the results of manual assessment. This method greatly increases the speed of coal assessment and reduces the subjectivity. The algorithm only works for preexisting images. The current study doesn’t address the creation of new images: the equipment required, sample preparation, and the specifics of microphotography.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Automated Identification of Coal’s Maceral Groups from Microphotographs Using Computer Vision and Convolutional Neural Networks

  • 期刊论文

  • 发表日期：

    2025年8月1日

查看全文
  • M. V. Shishanov,
  • N. N. Dobryakova,
  • M. S. Luchkin,
  • A. S. Evtiugin,
  • I. S. Mezrin,
  • A. Yu. Bozhko

摘要

Abstract

A method is developed for automated identification of groups of macerals (vitrinite, inertinite, liptinite, and semivitrinite) and mineral inclusions from digital microphotographs of polished coal briquet surfaces. It is based on convolutional neural networks (CNN) and computer vision. For a test sample, the accuracy attained is 92.31%; that is comparable with the results of manual assessment. This method greatly increases the speed of coal assessment and reduces the subjectivity. The algorithm only works for preexisting images. The current study doesn’t address the creation of new images: the equipment required, sample preparation, and the specifics of microphotography.