MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope Petrography

<p dir="ltr">Morphological characterization of microcrystalline rock textures typically relies upon the visual interpretation and manual measurement of scanning electron microscopy (SEM) imagery: a practice fraught with subjectivity, inefficiency, sampling bias, and data loss. We int...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Mohammed Yaqoob (344668) (author)
مؤلفون آخرون: Mohammed Yusuf Ansari (22047911) (author), Mohammed Ishaq (22302736) (author), Issac Sujay Anand John Jayachandran (22457725) (author), Mohammed S. Hashim (21586770) (author), Thomas Daniel Seers (21797324) (author)
منشور في: 2025
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1864513534721785856
author Mohammed Yaqoob (344668)
author2 Mohammed Yusuf Ansari (22047911)
Mohammed Ishaq (22302736)
Issac Sujay Anand John Jayachandran (22457725)
Mohammed S. Hashim (21586770)
Thomas Daniel Seers (21797324)
author2_role author
author
author
author
author
author_facet Mohammed Yaqoob (344668)
Mohammed Yusuf Ansari (22047911)
Mohammed Ishaq (22302736)
Issac Sujay Anand John Jayachandran (22457725)
Mohammed S. Hashim (21586770)
Thomas Daniel Seers (21797324)
author_role author
dc.creator.none.fl_str_mv Mohammed Yaqoob (344668)
Mohammed Yusuf Ansari (22047911)
Mohammed Ishaq (22302736)
Issac Sujay Anand John Jayachandran (22457725)
Mohammed S. Hashim (21586770)
Thomas Daniel Seers (21797324)
dc.date.none.fl_str_mv 2025-04-01T03:00:00Z
dc.identifier.none.fl_str_mv 10.1109/access.2025.3552626
dc.relation.none.fl_str_mv https://figshare.com/articles/journal_contribution/MicroCrystalNet_An_Efficient_and_Explainable_Convolutional_Neural_Network_for_Microcrystal_Classification_Using_Scanning_Electron_Microscope_Petrography/30393235
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Engineering
Resources engineering and extractive metallurgy
Information and computing sciences
Artificial intelligence
Machine learning
Carbonate characterization
classification
deep learning
petrography
segmentation
SEM imaging
materials science
Rocks
Scanning electron microscopy
Crystals
Pipelines
Feature extraction
Photomicrography
Visualization
Media
Surface topography
Surface morphology
dc.title.none.fl_str_mv MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope Petrography
dc.type.none.fl_str_mv Text
Journal contribution
info:eu-repo/semantics/publishedVersion
text
contribution to journal
description <p dir="ltr">Morphological characterization of microcrystalline rock textures typically relies upon the visual interpretation and manual measurement of scanning electron microscopy (SEM) imagery: a practice fraught with subjectivity, inefficiency, sampling bias, and data loss. We introduce a state-of-the-art computer vision pipeline, built on deep learning architectures, for segmenting and classifying individual microcrystals from SEM images. Initially applied to low-Mg calcite carbonate rocks, instance segmentation is achieved using a custom-tuned version of Meta’s Segment Anything Model (SAM). To train and test the classifier, we utilized 48 SEM images of diverse carbonate microtextures composed of Low-Mg calcite from studies performed worldwide. Each individual microcrystal (1852 in total) was labelled according to a bipartite classification scheme, encompassing both crystal shape (rhombic, polyhedral, amorphous, and spherical), and degree of crystal facet definition (euhedral to subhedral, anhedral), with a total of four distinct classes. MicroCrystalNet: our proposed classification model employs a convolutional neural network architecture, incorporating advanced feature map processing (feature normalization, dimensionality reduction, and sparse feature selection), integrated within a novel Normalized Sparse Reduction block. Performance metrics reveal excellent average precision scores (AP = 0.93-0.98) and Area Under Receiver-Operator Curve values (AUC = 0.95-0.99) across all classes, with visual comparison to manual ground truth images demonstrating powerful inter-class discriminatory power, even in the presence of occlusions.</p><h2>Other Information</h2><p dir="ltr">Published in: IEEE Access<br>License: <a href="https://creativecommons.org/licenses/by/4.0/deed.en" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1109/access.2025.3552626" target="_blank">https://dx.doi.org/10.1109/access.2025.3552626</a></p>
eu_rights_str_mv openAccess
id Manara2_28ef981213b4fe6789d14c9f225297fb
identifier_str_mv 10.1109/access.2025.3552626
network_acronym_str Manara2
network_name_str Manara2
oai_identifier_str oai:figshare.com:article/30393235
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope PetrographyMohammed Yaqoob (344668)Mohammed Yusuf Ansari (22047911)Mohammed Ishaq (22302736)Issac Sujay Anand John Jayachandran (22457725)Mohammed S. Hashim (21586770)Thomas Daniel Seers (21797324)EngineeringResources engineering and extractive metallurgyInformation and computing sciencesArtificial intelligenceMachine learningCarbonate characterizationclassificationdeep learningpetrographysegmentationSEM imagingmaterials scienceRocksScanning electron microscopyCrystalsPipelinesFeature extractionPhotomicrographyVisualizationMediaSurface topographySurface morphology<p dir="ltr">Morphological characterization of microcrystalline rock textures typically relies upon the visual interpretation and manual measurement of scanning electron microscopy (SEM) imagery: a practice fraught with subjectivity, inefficiency, sampling bias, and data loss. We introduce a state-of-the-art computer vision pipeline, built on deep learning architectures, for segmenting and classifying individual microcrystals from SEM images. Initially applied to low-Mg calcite carbonate rocks, instance segmentation is achieved using a custom-tuned version of Meta’s Segment Anything Model (SAM). To train and test the classifier, we utilized 48 SEM images of diverse carbonate microtextures composed of Low-Mg calcite from studies performed worldwide. Each individual microcrystal (1852 in total) was labelled according to a bipartite classification scheme, encompassing both crystal shape (rhombic, polyhedral, amorphous, and spherical), and degree of crystal facet definition (euhedral to subhedral, anhedral), with a total of four distinct classes. MicroCrystalNet: our proposed classification model employs a convolutional neural network architecture, incorporating advanced feature map processing (feature normalization, dimensionality reduction, and sparse feature selection), integrated within a novel Normalized Sparse Reduction block. Performance metrics reveal excellent average precision scores (AP = 0.93-0.98) and Area Under Receiver-Operator Curve values (AUC = 0.95-0.99) across all classes, with visual comparison to manual ground truth images demonstrating powerful inter-class discriminatory power, even in the presence of occlusions.</p><h2>Other Information</h2><p dir="ltr">Published in: IEEE Access<br>License: <a href="https://creativecommons.org/licenses/by/4.0/deed.en" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1109/access.2025.3552626" target="_blank">https://dx.doi.org/10.1109/access.2025.3552626</a></p>2025-04-01T03:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.1109/access.2025.3552626https://figshare.com/articles/journal_contribution/MicroCrystalNet_An_Efficient_and_Explainable_Convolutional_Neural_Network_for_Microcrystal_Classification_Using_Scanning_Electron_Microscope_Petrography/30393235CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/303932352025-04-01T03:00:00Z
spellingShingle MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope Petrography
Mohammed Yaqoob (344668)
Engineering
Resources engineering and extractive metallurgy
Information and computing sciences
Artificial intelligence
Machine learning
Carbonate characterization
classification
deep learning
petrography
segmentation
SEM imaging
materials science
Rocks
Scanning electron microscopy
Crystals
Pipelines
Feature extraction
Photomicrography
Visualization
Media
Surface topography
Surface morphology
status_str publishedVersion
title MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope Petrography
title_full MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope Petrography
title_fullStr MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope Petrography
title_full_unstemmed MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope Petrography
title_short MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope Petrography
title_sort MicroCrystalNet: An Efficient and Explainable Convolutional Neural Network for Microcrystal Classification Using Scanning Electron Microscope Petrography
topic Engineering
Resources engineering and extractive metallurgy
Information and computing sciences
Artificial intelligence
Machine learning
Carbonate characterization
classification
deep learning
petrography
segmentation
SEM imaging
materials science
Rocks
Scanning electron microscopy
Crystals
Pipelines
Feature extraction
Photomicrography
Visualization
Media
Surface topography
Surface morphology