An embedded multispectral imaging system to characterize vineyard canopy development and its sanitary state evolution

Abstract

Multispectral imaging systems are widely used in remote sensing and applied to viticulture context for the canopy characterization. Until now, this technique was scarcely used in proximal sensing to characterize vine canopy. Yet the results of field tests led in fixed position have revealed its capacity to estimate the leaf area. The aim of this project is to assess the opportunity to use multispectral imaging system as an embedded sensor for vine canopy characterization.

To this end, a multispectral camera acquiring visible and near-infrared images and a Greenseeker RT-100 device providing an NDVI (Normalized Difference Vegetation Index), were installed on a track laying tractor. It was equipped with a centimetric GPS device for data georeferencing. Image acquisition was done outdoor (using natural sunlight) in the trellis plan. The experimental field was composed of three grape varieties (Pinot Noir, Chardonnay and Meunier) that were planted according to the latin square pattern. Several datasets were collected in 2013. Theycorresponded to different phenological stages of the vine, from the berry formation to a stage close to harvest date.

This work will present the advantage of imagery technique specifically on the characterization of vegetation in the grape area. We will present the evolution of the vegetation in this area and compared it to ones considered in the general trellis high. Three grapevine varieties and two different training systems were e considered in this experiment : Chablis pruned and Cordon (or Simple- Guyot) . Data analysis shown multispectral imaging is an accurate and versatile technique to characterise vine canopy especially with NDVI applications.