Exploring the Potential of Thermal Point Clouds to Assess Crop Water Stress Within Precision Viticulture

Abstract

The Crop Water Stress Index (CWSI) has emerged as a valuable tool in Precision Viticulture (PV) for assessing crop water stress over vast areas. This is particularly significant considering the increasing demand for water and limited water resources. Maximizing water efficiency and crop yield through irrigation scheduling is therefore of utmost importance. While various techniques have been employed to improve crop water stress analysis, the use of thermal point clouds remain relatively unexplored. This thesis aims to enhance crop water stress analysis methodologies within PV through the integration of remote sensing, thermal imaging, and point cloud technologies. The study focuses on evaluating the viability of using thermal point clouds to generate 3D point clouds with CWSI values and 2D CWSI orthomosaics. Comparative analysis of these data models determines the appropriateness of employing thermal point clouds for CWSI calculation within PV. Additionally, the research explores the influence of different flight configurations (nadir, oblique, and their combination) on CWSI outcomes, seeking to identify the most effective workflow for CWSI calculation using point clouds. The findings exhibit promising potential, showing that thermal point clouds can be effectively employed to generate CWSI point clouds. The research findings indicate that point clouds offer a more comprehensive representation of the canopy compared to orthomosaics, thus providing more detailed information. Volume calculations show that the combined workflow yields the most accurate results in terms of geometric representation (R2 = 0.72), followed by the nadir flight (R2 = 0.68), and finally the oblique flight (R2 = 0.54). This outcome indicates that the combined workflow is the optimal approach for CWSI calculations utilizing point clouds.