APPLICATION OF NEURAL NETWORKS IN METALLOGRAPHY
ENRBTY
DOI:
https://doi.org/10.25712/ASTU.2072-8921.2024.02.023Abstract
The article considers the possibility of using artificial neural networks to solve the problem of classification of steel grades by microstructure image. To solve the recognition problem, the structure of a convolutional neural network is selected. Images for creating Data sata are pre-pared using a metallographic microscope, the environment for writing a neural network is made in Python Google Colaboratory. The convolutional neural network was trained at 96.67%. High results of training in the classification of images of microstructures of carbon steels indicate the prospects of using neural networks in metallographic studies
References
Аристов Г.В., Клюев А.В. Распознавание и классификация микроструктуры металлов и сплавов с использованием глубоких нейронных сетей // GraphiCon 2017 : тр. 27-й Между-нар. конф. по компьютерной графике и машинному зрению.Пермь, 24–28 сент. 2017 г. / АНО науч. общ-во Графикон, Перм. гос. нац. исслед. ун-т. Пермь: Изд-воЦентрПГНИУ. 2017. С. 180–183.
Kovun V.A. On development of models and al-gorithms for automated metallographic measurement of visible metal slice grain sizes / V.A. Kovun, I.L. Kashirina // Journal of Physics: Conference Series, 2020.
КовунВ.А., КаширинаИ.Л. Использование нейронной сети W-Net в металлографическом анализе образца стали / информатика: проблемы, методы, технологии, Материалы XXI Международной научно-методической конференции. Воронеж, 2021 / Изд-во «Вэлборн» (Воронеж). С. 760–767
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2024 Vladimir I. Mosorov, Natalya B. Khaptakhaeva, Konstantin S. Korobkov
This work is licensed under a Creative Commons Attribution 4.0 International License.
.
This work is licensed under a 
