Solving the problem of recognition and segmentation of images of natural resources deposits based on an ensemble of neural networks

A.I. Komarov, V.V. Kholmogorov, E.M. Zabotkina, A.I. Vakulenko, D.I. Chugunova


Exploration of mineral deposits is a complex task, for the effective solution of which it is necessary to obtain the maximum possible amount of relevant information about the object under study. The article considers a possible solution to the problem of exploration of mineral deposits by using the tools of artificial neural networks. The main idea in building the architecture in this work was the ability to combine the segmentation of mineral deposits on the geological map of the region and their classification by type and volume. An ensemble of clustering algorithms has been developed to solve the segmentation problem. This ensemble provides a single vector of cluster labels describing the affiliation of geo-system points to certain clusters. To solve the classification problem, it was decided to use an ensemble of DeiT models that provide high-quality classification of hyperspectral images in the task of searching for ore deposits. The use of an ensemble of DeiT models has advantages over the use of individual DeiT models, in particular, this approach reduces the effect of retraining and improves the quality of classification.

Full Text:

PDF (Russian)


Modern methods of analytical data processing / N. S. Abramov, D. A. Makarov, A. A. Talalaev, V. P. Ralenko// Software systems: theory and application/─2018.─ 4(39).─ pp.417-442

A review of the past 5 years of international advances in multi- and hyperspectral satellite data application and processing techniques in geological research/ Smirnova I.O., A. Kirsanov.A.1, Kamyshnikova N.V.//Modern problems of remote sensing of the Earth from space/-2020.-1.-gr.9-27

Possibilistic-fuzzy segmentation of earth surface images by means of genetic algorithms and artificial neural networks /L.A.Demidova, N.I.Nesterov, R.V.Tishkin//Scientific and technical bulletin of SPbSPU/─2014.─3(198) ─ pp. 37-48

Segmentation of objects according to hyperspectral survey of the Earth using artificial intelligence methods / Demidova L.A., V. Eremeev.V., Myatov G.N., Tishkin R.V., Yudakov A.A.// Digital Signal Processing/-2013.-4.-gr.32-36

ViT-DeiT: An ensemble model for the classification of histopathological images of breast cancer / Amira Alotaibi, Tariq Alafif, Faris Alkhilaivi, Yasir Alatavi, Hassan Altobaiti, Abdulmajid Alrefai, Youssef M Hawsawi, Tin Nguyen: // [Electronic resource]. URL: November 1, 2022 (Accessed 04/18/2023)

Ensembles of information‐efficient vision converters as a new paradigm of automated classification in ecology / S. P. Kyatanahalli1, T. Hardeman, M. Reyes, E. Merz, T. Bulas, P. Brun, F. Pomati, M. Bayti‐Jesy // [Electronic resource]. URL: / / Published: November 03, 2022 (Accessed 04/18/2023)

Sokolov S. At McKinsey: how to save $ 370 billion a year in mining due to digital technologies: // [Electronic resource]. URL: (Date of announcement 15.04.2023]

Jeffrey E. Hinton, Oriol Vinyals, J. Dean Distilling the Knowledge in a Neural Network: // [Electronic resource]. URL: Published on March 9, 2015 (Accessed 04/18/2023)


  • There are currently no refbacks.

Abava  Кибербезопасность MoNeTec 2024

ISSN: 2307-8162