Introducing on-the-go sensing rover for vines canopy abiotic stressors detection

Abstract

In this paper, we proposed a new unmanned terrestrial vehicle (UTV) for the detection of vine canopy abiotic stress. The developed UTV is a customized version of an open source, open hardware robotic platform called MARRtino. The developed UTV has been equipped with a series of different cameras (RGB, FLIR and multispectral) that automatically acquire images of vineyard rows. Acquired images have been post processed applying stitching algorithms and generating the ortomosaic of the vine canopy. Structure from Motion (SFM) image technique has been then used to generate a 3D point cloud of the vineyard rows. Generated point clouds are very useful for canopy reconstruction, phenotyping and for information extraction that can be correlated to the management system of the vineyard. The size and volume of canopy can be converted into canopy indicators and linked to vegetation indices. These indices will be correlated with the ground-truth for defining the site-specific map in the vineyard and implementing a local monitoring for carry out the prescription map.