Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks

Assiya Sarinova; Leila Rzayeva; Noyan Tendikov; Ibraheem Shayea

doi:10.1109/WINCOM59760.2023.10323012

Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks

Assiya Sarinova^*, Leila Rzayeva, Noyan Tendikov, Ibraheem Shayea

^*Corresponding author for this work

Faculty of Electrical and Electronics

Astana IT University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Citations (Scopus)

Abstract

This paper introduces a simple implementation of three versions (large, medium, and small) of terrain multi-classification models using Fully Convolutional Neural Networks (FCNNs) for imagery data. The proposed methodology involves labeled and unlabeled data collection from European Space Agency (ESA) WorldCover and Sentinel-2 MultiSpectral Instrument (MSI) on the Google Earth Engine, compressing datasets into Tensorflow records format with 9 diverse terrain types, and handling Google Cloud training computations. There were prepared different dataset portions of 10 megabytes, 200 megabytes, and around a gigabyte files. The experimental results demonstrate the effectiveness of the CNN-based approach, achieving a tolerable 71% accuracy of the Terrain Classification Model (TCM) and robust classification performance. The simplicity and efficiency of the proposed method make it suitable for real-world applications requiring reliable and fast terrain classification.

Original language	English
Title of host publication	Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023
Editors	Khalil Ibrahimi, Mohamed El Kamili, Abdellatif Kobbane, Ibraheem Shayea
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9798350329674
DOIs	https://doi.org/10.1109/WINCOM59760.2023.10323012
Publication status	Published - 2023
Event	10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023 - Istanbul, Turkey Duration: 26 Oct 2023 → 28 Oct 2023

Publication series

Name	Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023

Conference

Conference	10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023
Country/Territory	Turkey
City	Istanbul
Period	26/10/23 → 28/10/23

Bibliographical note

Publisher Copyright:
© 2023 IEEE.

Keywords

Aerospace images
Deep Learning
Google Cloud Computing
Machine Learning
Satellite data
Terrain Classification

Access to Document

10.1109/WINCOM59760.2023.10323012

Cite this

Sarinova, A., Rzayeva, L., Tendikov, N., & Shayea, I. (2023). Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks. In K. Ibrahimi, M. El Kamili, A. Kobbane, & I. Shayea (Eds.), Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023 (Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/WINCOM59760.2023.10323012

Sarinova, Assiya ; Rzayeva, Leila ; Tendikov, Noyan et al. / Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks. Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023. editor / Khalil Ibrahimi ; Mohamed El Kamili ; Abdellatif Kobbane ; Ibraheem Shayea. Institute of Electrical and Electronics Engineers Inc., 2023. (Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023).

@inproceedings{f58a97d54a9740078f89047b94422262,

title = "Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks",

abstract = "This paper introduces a simple implementation of three versions (large, medium, and small) of terrain multi-classification models using Fully Convolutional Neural Networks (FCNNs) for imagery data. The proposed methodology involves labeled and unlabeled data collection from European Space Agency (ESA) WorldCover and Sentinel-2 MultiSpectral Instrument (MSI) on the Google Earth Engine, compressing datasets into Tensorflow records format with 9 diverse terrain types, and handling Google Cloud training computations. There were prepared different dataset portions of 10 megabytes, 200 megabytes, and around a gigabyte files. The experimental results demonstrate the effectiveness of the CNN-based approach, achieving a tolerable 71\% accuracy of the Terrain Classification Model (TCM) and robust classification performance. The simplicity and efficiency of the proposed method make it suitable for real-world applications requiring reliable and fast terrain classification.",

keywords = "Aerospace images, Deep Learning, Google Cloud Computing, Machine Learning, Satellite data, Terrain Classification",

author = "Assiya Sarinova and Leila Rzayeva and Noyan Tendikov and Ibraheem Shayea",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023 ; Conference date: 26-10-2023 Through 28-10-2023",

year = "2023",

doi = "10.1109/WINCOM59760.2023.10323012",

language = "English",

series = "Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

editor = "Khalil Ibrahimi and \{El Kamili\}, Mohamed and Abdellatif Kobbane and Ibraheem Shayea",

booktitle = "Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023",

address = "United States",

}

Sarinova, A, Rzayeva, L, Tendikov, N & Shayea, I 2023, Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks. in K Ibrahimi, M El Kamili, A Kobbane & I Shayea (eds), Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023. Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023, Institute of Electrical and Electronics Engineers Inc., 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023, Istanbul, Turkey, 26/10/23. https://doi.org/10.1109/WINCOM59760.2023.10323012

Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks. / Sarinova, Assiya; Rzayeva, Leila; Tendikov, Noyan et al.
Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023. ed. / Khalil Ibrahimi; Mohamed El Kamili; Abdellatif Kobbane; Ibraheem Shayea. Institute of Electrical and Electronics Engineers Inc., 2023. (Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks

AU - Sarinova, Assiya

AU - Rzayeva, Leila

AU - Tendikov, Noyan

AU - Shayea, Ibraheem

PY - 2023

Y1 - 2023

N2 - This paper introduces a simple implementation of three versions (large, medium, and small) of terrain multi-classification models using Fully Convolutional Neural Networks (FCNNs) for imagery data. The proposed methodology involves labeled and unlabeled data collection from European Space Agency (ESA) WorldCover and Sentinel-2 MultiSpectral Instrument (MSI) on the Google Earth Engine, compressing datasets into Tensorflow records format with 9 diverse terrain types, and handling Google Cloud training computations. There were prepared different dataset portions of 10 megabytes, 200 megabytes, and around a gigabyte files. The experimental results demonstrate the effectiveness of the CNN-based approach, achieving a tolerable 71% accuracy of the Terrain Classification Model (TCM) and robust classification performance. The simplicity and efficiency of the proposed method make it suitable for real-world applications requiring reliable and fast terrain classification.

AB - This paper introduces a simple implementation of three versions (large, medium, and small) of terrain multi-classification models using Fully Convolutional Neural Networks (FCNNs) for imagery data. The proposed methodology involves labeled and unlabeled data collection from European Space Agency (ESA) WorldCover and Sentinel-2 MultiSpectral Instrument (MSI) on the Google Earth Engine, compressing datasets into Tensorflow records format with 9 diverse terrain types, and handling Google Cloud training computations. There were prepared different dataset portions of 10 megabytes, 200 megabytes, and around a gigabyte files. The experimental results demonstrate the effectiveness of the CNN-based approach, achieving a tolerable 71% accuracy of the Terrain Classification Model (TCM) and robust classification performance. The simplicity and efficiency of the proposed method make it suitable for real-world applications requiring reliable and fast terrain classification.

KW - Aerospace images

KW - Deep Learning

KW - Google Cloud Computing

KW - Machine Learning

KW - Satellite data

KW - Terrain Classification

UR - https://www.scopus.com/pages/publications/85179519739

U2 - 10.1109/WINCOM59760.2023.10323012

DO - 10.1109/WINCOM59760.2023.10323012

M3 - Conference contribution

AN - SCOPUS:85179519739

T3 - Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023

BT - Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023

A2 - Ibrahimi, Khalil

A2 - El Kamili, Mohamed

A2 - Kobbane, Abdellatif

A2 - Shayea, Ibraheem

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023

Y2 - 26 October 2023 through 28 October 2023

ER -

Sarinova A, Rzayeva L, Tendikov N, Shayea I. Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks. In Ibrahimi K, El Kamili M, Kobbane A, Shayea I, editors, Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023. Institute of Electrical and Electronics Engineers Inc. 2023. (Proceedings - 10th International Conference on Wireless Networks and Mobile Communications, WINCOM 2023). doi: 10.1109/WINCOM59760.2023.10323012

Simple Implementation of Terrain Classification Models via Fully Convolutional Neural Networks

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this