Data Complexity-Oriented Classification of Multispectral Remote Sensing Imagery via Machine and Deep Learning Approaches

Berrin Islek, Hamza Erol

2025

Abstract

In this study, the land cover of an agricultural region was classified at a field level using multispectral satellite imagery. The primary objective of the study was to evaluate different classification methods in terms of data complexity, computational complexity, and information complexity. The data labelling process was performed using hierarchical clustering, making the groups in the data more meaningful. A separate clustering tree structure was created for each feature, and data complexity was analysed using parameters such as level, number of families, and number of children. Object-oriented approaches were adopted in the classification phase, employing Deep Neural Networks, Random Forest, and Support Vector Machines. The performance of these methods was examined not only in terms of accuracy but also in terms of evaluation metrics such as F1-score, recall, and precision. The results demonstrate the classification capabilities of the methods in a comprehensive manner and provide important clues about which approach is more suitable in different scenarios. Furthermore, the methods were compared in terms of computational costs and processing times, and a comprehensive evaluation was conducted regarding the classification of agricultural regions using remotely sensed data.

Paper Citation

in Harvard Style

Islek B. and Erol H. (2025). Data Complexity-Oriented Classification of Multispectral Remote Sensing Imagery via Machine and Deep Learning Approaches. In Proceedings of the 2nd International Conference on Advances in Electrical, Electronics, Energy, and Computer Sciences - Volume 1: ICEEECS; ISBN 978-989-758-783-2, SciTePress, pages 257-265. DOI: 10.5220/0014313500004848

in Bibtex Style

@conference{iceeecs25,
author={Berrin Islek and Hamza Erol},
title={Data Complexity-Oriented Classification of Multispectral Remote Sensing Imagery via Machine and Deep Learning Approaches},
booktitle={Proceedings of the 2nd International Conference on Advances in Electrical, Electronics, Energy, and Computer Sciences - Volume 1: ICEEECS},
year={2025},
pages={257-265},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0014313500004848},
isbn={978-989-758-783-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Advances in Electrical, Electronics, Energy, and Computer Sciences - Volume 1: ICEEECS
TI - Data Complexity-Oriented Classification of Multispectral Remote Sensing Imagery via Machine and Deep Learning Approaches
SN - 978-989-758-783-2
AU - Islek B.
AU - Erol H.
PY - 2025
SP - 257
EP - 265
DO - 10.5220/0014313500004848
PB - SciTePress