Method of constructing artifi cial neural networks for the identifi cation of special purpose structures by classifi cation and country of origin based on electron microscopic analysis of their fragments

Authors
DOI:
https://doi.org/10.34169/2414-0651.2024.1(41).128-140
Keywords:
Array, Array, Array, Array, Array
Abstract

The article describes a methodology that employs visual analysis of metal microstructures based on AI for their efficient classification. This approach is grounded on a dataset composed of microscopic images depicting various metal classes, each showcasing distinct visual features such as grain size, shape, phase distribution, inclusions, and signs of mechanical processing. The use of convolutional neural networks (CNNs) for analyzing these features has significantly improved the identification and classification process, demonstrating the potential of AI in materials science.

Experimental results have confirmed the effectiveness of using CNNs to accurately distinguish between two and three metal classes. One notable observation was the positive impact of increasing image sizes on the classification process. This fact suggests that higher-quality images can mitigate the limitations of a small dataset size, thereby enriching the analysis.

In addition to convolutional neural networks, the paper describes the features of involving the large language model GPT-4v in the classification of metals and demonstrates its effectiveness.

Author Biographies
  1. Igor Chepkov , Central Scientific Research Institute of Armament and Military Equipment of the Armed Forces of Ukraine

    Doctor of Technical Sciences, Professor

  2. Vadym Slyusar, Central Scientific Research Institute of Armament and Military Equipment of the Armed Forces of Ukraine

    Doctor of Technical Sciences, Professor

  3. Andrii Kuchinskiy, Central Scientific Research Institute of Armament and Military Equipment of the Armed Forces of Ukraine

    Candidate of Technical Sciences, Senior Scientist

References

Слюсар В.І. Архітектурно-математичні основи удосконалення нейронних мереж з класифікації зображень. Штучний інтелект. 2022. № 1. С. 127—138.

https://doi.org/10.15407/jai2022.01.245

Slyusar, V., Protsenko, M., Chernukha, A., Gornostal, S., Rudakov, S., Shevchenko, S., Chernikov, O., Kolpachenko, N., Timofeyev, V. & Artiukh, R. (2021). Construction of an advanced method for recognizing monitored objects by a convolutional neural network using a discrete wavelet transform. Eastern-European J. of Enterprise Technologies. No 4(9(112)). Pp. 65—77.

https://doi.org/10.15587/1729-4061.2021.238601.

Слюсар В.І., Проценко М.М., Докучаєв О.В. Методика побудови нейронних мереж для ідентифікації озброєння та військової техніки. Озброєння та військова техніка. 2022. № 2(34). С. 99—107. https://doi.org/10.34169/2414-0651.2022.2(34).99-107

Vadym Slyusar, Mykhailo Protsenko & Mariia Bondarchuk. (2022). Methodology for Armaments Identification Using a Neural Network. Problems of Infocommunications. Science and Technology: 2022 Intern. Scient.-Pract. Conf. PIC S&T. 5 p.

https://doi.org/10.1109/PICST57299.2022.10238568.

Slyusar, V.I. &, Sliusar, I.I. (2021). Lions of Neural Networks Zoo, Neyromerezhni tehnologii ta yih zastosuvannia NMTIZ-2021: zb. nauk. pr. XX Mizhn. Nauk. konf. «Neiromerezhny tehnologii ta yih zastosuvannia NMTIZ-2021». Kramatorsk: DDMA. Pp. 129—133.

https://doi.org/10.13140/RG.2.2.17187.58405.

Qassim, H., Verma, A. & Feinzimer, D. (2018). Compressed residual-VGG16 CNN model for big data places image recognition. Computing and Communication Workshop and Conf. (CCWC) 2018 IEEE 8th Annual. Pp. 169—175.

M. Tan & Q. Le. (2019). EfficientNet: Rethinking model scaling for convolutional neural networks; in Proc. of Intern. Conf. on Machine Learning. Pp. 6105—6114.

Sandler, M., Howard, A. & Zhu, M. (2018) Mobilenetv2: Inverted Residuals and Linear Bottlenecks. Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018. Pp. 4510—4520.

https://doi.org/10.1109/CVPR.2018.00474.

Howard, A., Sandler, M. & Chu, G. (2019) Searching for Mobilenetv3. Proc. of the IEEE Intern. Conf. on Computer Vision, Seoul, 27 October − November 2019. Pp. 1314—1324.

https://doi.org/10.1109/ICCV.2019.00140.

Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, & Li Fei-Fei. (2009). Imagenet: A large-scale hierarchical image database. In 2009 IEEE conf. on computer vision and pattern recognition. IEEE. Pp. 248—255.

Chollet, F. (2017). Xception: Deep Learning with Depthwise Separable Convolutions. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR). Pp. 1800—1807. https://doi.org/10.1109/CVPR.2017.195.

Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J. & Wojna, Z. (2016). Rethinking the Inception Architecture for Computer Vision. In 2016 IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA. Pp. 2818—2826.

https://doi.org/10.1109/CVPR.2016.308.

GPT-4. Technical Report by OpenAI, 27 March 2023. Available at: https://arxiv.org/pdf/2303.08774v3.pdf.

Liu, H., Li, C., Wu, Q. & Lee, Y.J. (2023). Visual Instruction Tuning. 19 p. Available at: https://arxiv.org/abs/2304.08485.

Cover Image
Downloads
Published
2024-03-31
Issue
Vol. 41 No. 1 (2024): Weapons and military equipment
Section
INFORMATION SYSTEMS
License

Copyright (c) 2024 Ігор Чепков,Вадим Слюсар,Андрій Кучинський

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

How to Cite

Method of constructing artifi cial neural networks for the identifi cation of special purpose structures by classifi cation and country of origin based on electron microscopic analysis of their fragments. (2024). Weapons and Military Equipment, 41(1), 128-140. https://doi.org/10.34169/2414-0651.2024.1(41).128-140

Most read articles by the same author(s)

<< < 1 2 3 4 5