Accurate Detection of Land Use and Land Cover Changes in Kabul City Using Advanced Machine Learning Algorithms Applied to Pan-Sharpened Satellite Imagery (2015–2023)
Kabul Polytechnic University International Journal of Engineering & Technology · 22.08.2026 · 6 (1) · 121-134
DOI: 10.68924/21yze196 ↗Kurz gefasst
Fachartikel in der Zeitschrift Kabul Polytechnic University International Journal of Engineering & Technology, online erschienen am 22.08.2026. Verfasst von MOHAMMAD KARIM SIRAT und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Efficient implementation of remote sensing image classification plays a crucial role in supporting the extraction of spatial and temporal information for land use and land cover (LU/LC) analysis. Mapping LU/LC changes allows the examination of various social, economic, and environmental effects on the Earth’s surface. This study aims to detect LU/LC changes in Kabul using an effective classification approach. Two classification techniques were evaluated: the Maximum Likelihood Classification (MLC) method and the Support Vector Machine (SVM) method, both implemented and compared using ENVI 5.6 software. To enhance the quality of the LU/LC map, the performance of the pan-sharpening algorithm was improved, increasing the spatial resolution of the satellite images from 30 meters to 15 meters. After classification, morphological operations with a 7×7 kernel were applied to reduce noise. LU/LC changes were analyzed using SVM for the years 2015 and 2023, achieving overall accuracies of 96.09% and 98.09% with Kappa coefficients of 0.92 and 0.94, respectively. The MLC method yielded overall accuracies of 94.03% and 97.34%, with Kappa coefficients of 0.88 and 0.91, respectively. Using the difference method on the SVM-classified images, the extent and pattern of LU/LC changes were assessed, and a detailed LU/LC change map for Kabul was generated. Results showed that between 2015 and 2023, urban land use increased by 7.31%, while vegetation cover, soil cover, and water bodies decreased by 1.18%, 6.11%, and 0.16%, respectively. The eight-year trend indicates that urban expansion in Kabul reflects broader patterns of social and economic development. This study offers new insights into detecting potential land use changes and supports the formulation of effective urban development strategies for future planning.
Zitieren
APA
SIRAT, M. K., SULTANI, M. Y., & EBRAHIMI, S. A. (2026). Accurate Detection of Land Use and Land Cover Changes in Kabul City Using Advanced Machine Learning Algorithms Applied to Pan-Sharpened Satellite Imagery (2015–2023). Kabul Polytechnic University International Journal of Engineering & Technology, 6(1), 121-134. https://doi.org/10.68924/21yze196
Vancouver
SIRAT MK, SULTANI MY, EBRAHIMI SA. Accurate Detection of Land Use and Land Cover Changes in Kabul City Using Advanced Machine Learning Algorithms Applied to Pan-Sharpened Satellite Imagery (2015–2023). Kabul Polytechnic University International Journal of Engineering & Technology. 2026;6(1):121-134. doi:10.68924/21yze196
BibTeX
@article{SIRAT2026,
author = {SIRAT, MOHAMMAD KARIM and SULTANI, MOHAMMAD YAQOOB and EBRAHIMI, SHIR ALI},
title = {Accurate Detection of Land Use and Land Cover Changes in Kabul City Using Advanced Machine Learning Algorithms Applied to Pan-Sharpened Satellite Imagery (2015–2023)},
journal = {Kabul Polytechnic University International Journal of Engineering & Technology},
year = {2026},
volume = {6},
number = {1},
pages = {121--134},
doi = {10.68924/21yze196}
}
EUVIMED