Funding tool: Basic research projects
Project-ID: FTI25-G-015
Project start: 01. Mai 2026
Project end: 30. April 2029
Runtime: 36 months / not yet started
Funding amount: € 360.000,00

Brief summary:
The project addresses the critical challenge of water conservation in agriculture, using viticulture as a case study due to its high value and climate change-related vulnerabilities. By integrating plant ecophysiology and computer science, the project aims to develop innovative, low-cost, and scalable methods for optimizing water use efficiency in vineyards. Central to the research is the use of consumer-grade smartphones for 3D canopy reconstruction via Structure-from-Motion (SfM), combined with multimodal imaging (thermal and multispectral) to estimate key plant traits such as leaf area, transpiration, and yield. Leaf area (LA) is a critical parameter that determines plant water evaporation. However, current methods for estimating leaf area are labor-intensive, often destructive, and lack the spatial and temporal resolution required for dynamic modeling. Moreover, when water deficits occur, the relationship between LA, plant transpiration, and climatic variables changes due to increased resistance in the water pathway through the stomata. By incorporating a multimodal component into the 3D reconstruction (e.g., infrared thermography and multispectral imagery), the project aims to account for changes in vine water use under drought conditions. Additionally, the project will investigate how grapevine canopy architecture, influenced by varying trimming heights, affects water consumption and productivity. The project outcomes include open-source algorithms, annotated datasets, and practical vineyard management recommendations, all of which will be disseminated through peer-reviewed publications, conferences, and public repositories. By bridging gaps in precision agriculture and plant phenotyping, the project contributes to sustainable viticulture and offers transferable methodologies for other crops and agricultural systems.

Keywords:
Viticulture, Evapotranspiration, 3D reconstruction, Artificial Intelligence, Machine Learning, Computer Vision