An Innovative Methodology to Be More Time-Efficient When Analyzing Data in Precision Viticulture

Abstract: Remote Sensing tools in Precision Viticulture to detect disease levels of plants implies big data-sets and time-consuming analysis of data. In this study, we used remote images collected by an Unmanned Aerial Vehicle equipped with a Micasense RedEdge.MXTM multispectral camera and a Terrestrial Laser Scanner (TLS). These tools gave us large amounts of data in a short period. The aim of this research was to develop a methodology that shortens the post-processing phase of data. The data sets were taken from a vineyard during two crop monitoring surveys in June and August 2018. The monitored vineyard is situated in the Piedmont region (Italy). As a first step in the data analysis procedures, we used photogrammetry approaches, as well as the Digital Terrestrial Model and the Digital Surface Model to detect the shape of single plants. The achieved results were then validated with the analysis obtained through the TLS. We then analyzed the reflectance of the canopy using open source software, to detect changes in the pixels about the reflectance curve between healthy plants and plants with disease. We expect that the proposed methodology will help us to be time-efficient and to detect condition of vegetative changes.

Read the paper

Authors: Rinaldi, Monica & Gallo, Raimondo & Daglio, Gabriele & Mazzetto, Fabrizio.

Associations: Faculty of Science and Technology – Fa.S.TFree University of Bozen – Bolzano, Italy