Application of Deep Learning in Petrographic Coal Images Segmentation

The study of the petrographic structure of medium- and high-rank coals is important from both a cognitive and a utilitarian point of view. The petrographic constituents and their individual characteristics and features are responsible for the properties of coal and the way it behaves in various tech...

Full description

Saved in:

Bibliographic Details
Main Authors:	Sebastian Iwaszenko, Leokadia Róg
Format:	article
Language:	EN
Published:	MDPI AG 2021
Subjects:	coal petrographic analysis macerals image analysis semantic segmentation convolutional neural networks Mineralogy QE351-399.2
Online Access:	https://doaj.org/article/90b906fa103447dfba663b97fbb55ce8
Tags:	Add Tag No Tags, Be the first to tag this record!

id	oai:doaj.org-article:90b906fa103447dfba663b97fbb55ce8
record_format	dspace
spelling	oai:doaj.org-article:90b906fa103447dfba663b97fbb55ce82021-11-25T18:26:42ZApplication of Deep Learning in Petrographic Coal Images Segmentation10.3390/min111112652075-163Xhttps://doaj.org/article/90b906fa103447dfba663b97fbb55ce82021-11-01T00:00:00Zhttps://www.mdpi.com/2075-163X/11/11/1265https://doaj.org/toc/2075-163XThe study of the petrographic structure of medium- and high-rank coals is important from both a cognitive and a utilitarian point of view. The petrographic constituents and their individual characteristics and features are responsible for the properties of coal and the way it behaves in various technological processes. This paper considers the application of convolutional neural networks for coal petrographic images segmentation. The U-Net-based model for segmentation was proposed. The network was trained to segment inertinite, liptinite, and vitrinite. The segmentations prepared manually by a domain expert were used as the ground truth. The results show that inertinite and vitrinite can be successfully segmented with minimal difference from the ground truth. The liptinite turned out to be much more difficult to segment. After usage of transfer learning, moderate results were obtained. Nevertheless, the application of the U-Net-based network for petrographic image segmentation was successful. The results are good enough to consider the method as a supporting tool for domain experts in everyday work.Sebastian IwaszenkoLeokadia RógMDPI AGarticlecoalpetrographic analysismaceralsimage analysissemantic segmentationconvolutional neural networksMineralogyQE351-399.2ENMinerals, Vol 11, Iss 1265, p 1265 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	coal petrographic analysis macerals image analysis semantic segmentation convolutional neural networks Mineralogy QE351-399.2
spellingShingle	coal petrographic analysis macerals image analysis semantic segmentation convolutional neural networks Mineralogy QE351-399.2 Sebastian Iwaszenko Leokadia Róg Application of Deep Learning in Petrographic Coal Images Segmentation
description	The study of the petrographic structure of medium- and high-rank coals is important from both a cognitive and a utilitarian point of view. The petrographic constituents and their individual characteristics and features are responsible for the properties of coal and the way it behaves in various technological processes. This paper considers the application of convolutional neural networks for coal petrographic images segmentation. The U-Net-based model for segmentation was proposed. The network was trained to segment inertinite, liptinite, and vitrinite. The segmentations prepared manually by a domain expert were used as the ground truth. The results show that inertinite and vitrinite can be successfully segmented with minimal difference from the ground truth. The liptinite turned out to be much more difficult to segment. After usage of transfer learning, moderate results were obtained. Nevertheless, the application of the U-Net-based network for petrographic image segmentation was successful. The results are good enough to consider the method as a supporting tool for domain experts in everyday work.
format	article
author	Sebastian Iwaszenko Leokadia Róg
author_facet	Sebastian Iwaszenko Leokadia Róg
author_sort	Sebastian Iwaszenko
title	Application of Deep Learning in Petrographic Coal Images Segmentation
title_short	Application of Deep Learning in Petrographic Coal Images Segmentation
title_full	Application of Deep Learning in Petrographic Coal Images Segmentation
title_fullStr	Application of Deep Learning in Petrographic Coal Images Segmentation
title_full_unstemmed	Application of Deep Learning in Petrographic Coal Images Segmentation
title_sort	application of deep learning in petrographic coal images segmentation
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/90b906fa103447dfba663b97fbb55ce8
work_keys_str_mv	AT sebastianiwaszenko applicationofdeeplearninginpetrographiccoalimagessegmentation AT leokadiarog applicationofdeeplearninginpetrographiccoalimagessegmentation
_version_	1718411145384034304

Application of Deep Learning in Petrographic Coal Images Segmentation

Similar Items