Projecte llegit
Títol: Feasibility study on the use of drone for soil moisture estimation
Estudiants que han llegit aquest projecte:
ZHANG, JIAQI (data lectura: 30-10-2025)- Cerca aquest projecte a Bibliotècnica
ZHANG, JIAQI (data lectura: 30-10-2025)Director/a: ESPONA DONÉS, MARGARIDA
Departament: MAT
Títol: Feasibility study on the use of drone for soil moisture estimation
Data inici oferta: 05-02-2025 Data finalització oferta: 05-10-2025
Estudis d'assignació del projecte:
- MU DRONS
| Tipus: Individual | |
| Lloc de realització: Fora UPC | |
| Supervisor/a extern: Eulàlia Parés | |
| Institució/Empresa: Centre Tecnològic de Telecomunicacions d | |
| Titulació del Director/a: PhD | |
| Paraules clau: | |
| UAV; Vegetation indices; Soil moisture; QGIS; Multispectral imagery | |
| Descripció del contingut i pla d'activitats: | |
| Overview (resum en anglès): | |
| Soil moisture is a key parameter in agriculture, hydrology, and environmental monitoring. Traditional measurement methods-such as ground-based sensors and satellite remote sensing-face limitations in terms of spatial resolution, cost, or operational convenience. With the continuous advancement of UAV technology and lightweight sensors, high-resolution and cost-effective soil moisture monitoring using UAVs has become feasible.
This study aims to explore the feasibility of estimating soil moisture using UAVs, with a focus on evaluating their technical feasibility and system performance. The research includes a literature review and methodological study, selection and assessment of UAV platforms and sensors, and processing and performance analysis based on open-source data. By using QGIS tools to process multispectral imagery and calculate various vegetation indices (such as NDVI, NDWI, SAVI, and PDI), the results indicate that UAVs possess strong potential for multispectral imaging and vegetation index analysis, which can be used to indirectly infer the spatial distribution of soil moisture. This study verifies the effectiveness of UAV remote sensing technology through the processing and analysis of two open datasets: the SiDroForest dataset (covering the Siberian boreal forest region) and the Côte d'Ivoire cocoa plantation dataset. Both datasets were acquired using UAVs equipped with RGB and multispectral cameras. Band calculations and vegetation index inversions-including NDVI, NDWI, SAVI, and PDI-were performed using the QGIS platform to generate multi-indicator visualization maps. The analysis results demonstrate that different vegetation indices perform well in bare soil identification, assessment of water stress in vegetated areas, and extraction of spatial heterogeneity features, providing an effective basis for the indirect estimation of soil moisture. |
|