A Study on the Classification of Shrubs and Grasses on the Tibetan Plateau Based on Unmanned Aerial Vehicle Multispectral Imagery DOI Creative Commons
Xiaoqiang Chen, Hui Deng,

Wenjiang Zhang

и другие.

Remote Sensing, Год журнала: 2024, Номер 16(21), С. 4106 - 4106

Опубликована: Ноя. 2, 2024

The ecosystem of the Qinghai–Tibet Plateau is highly fragile due to its unique geographical conditions, with vegetation playing a crucial role in maintaining ecological balance. Thus, accurately monitoring distribution plateau region paramount importance. This study employs UAV multispectral imagery combination four machine-learning models—Support Vector Machine (SVM), Decision Tree (DT), Extreme Gradient Boosting (XGBoost), and Random Forest (RF)—to investigate impact different features their combinations on fine classification shrubs grasses Plateau, including Salix psammophila, Populus simonii Carrière, Kobresia tibetica, pygmaea. results indicate that near-infrared spectral information can improve accuracy, improvements 5.21%, 1.65%, 6.64%, 5.03% for pygmaea, respectively. Feature selection effectively reduces redundant enhances model all models achieving best performance optimized feature set. Furthermore, RF performs set, an overall accuracy (OA) 95.32% kappa coefficient 0.94. provides important scientific support vegetation.

Язык: Английский

Mapping Urban Green Spaces in Indonesian Cities Using Remote Sensing Analysis DOI Creative Commons
Agustiyara Agustiyara, Dyah Mutiarin,

Achmad Nurmandi

и другие.

Urban Science, Год журнала: 2025, Номер 9(2), С. 23 - 23

Опубликована: Янв. 22, 2025

This study explores the dynamics of urban green spaces in five major Indonesian cities—Central Jakarta, Bandung, Yogyakarta, Surabaya, and Semarang—using Sentinel-2 satellite imagery vegetation indices, such as NDVI EVI. As areas expand become more densely populated, development activities have significantly altered spaces, necessitating comprehensive mapping through remote sensing technologies. The findings reveal significant variability space coverage among cities over three periods (2019–2020, 2021–2022, 2023–2024), ensuring that are up to date. demonstrates utility for detailed analysis, emphasizing its effectiveness identifying, quantifying, monitoring changes spaces. Integrating advanced techniques, EVI, offers a nuanced understanding their implications sustainable planning. Utilizing data within Google Earth Engine (GEE) framework represents contemporary innovative approach studies, particularly rapidly urbanizing environments. novelty this research lies method preserving enhancing infrastructure while supporting effective strategies growth.

Язык: Английский

Процитировано

1
Urban green space vegetation height modeling and intelligent classification based on UAV multi-spectral and oblique high-resolution images DOI Creative Commons
Ronghua Li,

Zhican Bai,

Chao Ye

и другие.

Urban forestry & urban greening, Год журнала: 2025, Номер unknown, С. 128785 - 128785

Опубликована: Март 1, 2025

Язык: Английский

Процитировано

0
A Study on the Classification of Shrubs and Grasses on the Tibetan Plateau Based on Unmanned Aerial Vehicle Multispectral Imagery DOI Creative Commons
Xiaoqiang Chen, Hui Deng,

Wenjiang Zhang

и другие.

Remote Sensing, Год журнала: 2024, Номер 16(21), С. 4106 - 4106

Опубликована: Ноя. 2, 2024

The ecosystem of the Qinghai–Tibet Plateau is highly fragile due to its unique geographical conditions, with vegetation playing a crucial role in maintaining ecological balance. Thus, accurately monitoring distribution plateau region paramount importance. This study employs UAV multispectral imagery combination four machine-learning models—Support Vector Machine (SVM), Decision Tree (DT), Extreme Gradient Boosting (XGBoost), and Random Forest (RF)—to investigate impact different features their combinations on fine classification shrubs grasses Plateau, including Salix psammophila, Populus simonii Carrière, Kobresia tibetica, pygmaea. results indicate that near-infrared spectral information can improve accuracy, improvements 5.21%, 1.65%, 6.64%, 5.03% for pygmaea, respectively. Feature selection effectively reduces redundant enhances model all models achieving best performance optimized feature set. Furthermore, RF performs set, an overall accuracy (OA) 95.32% kappa coefficient 0.94. provides important scientific support vegetation.

Язык: Английский

Процитировано

0