Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
6540447 | Computers and Electronics in Agriculture | 2016 | 6 Pages |
Abstract
The vigor and the vegetative expression of vines are parameters of great interest in viticulture since they are indicative of the entire guidance system state. Traditionally these parameters are estimated and obtained using a destructive method by cutting the sticks, weighing them and finally obtain the medium weight of pruning sticks. This paper presents a non-destructive method that successfully determines the vine vigor based on a digital image processing. The method consists of acquiring a digital image of vines and applying a software to those images to determine vigor and vegetative expression. This application can be divided in two independent phases: the first one consists in a digital image processing, that isolate the vine sticks and determines the mean area per stick of the vine; the second one determines the vine vigor and vegetative expression based on a polynomial relationship between mean area and mean weight of stick. The software application was tested on thirty vines (Vitis vinifera cv. Touriga Nacional) characterized by low, medium and high vigor and selected during a three-year study. Since the mean weight of stick obtained using the mathematical expression and the values calculated using the real sticks present a 0.998 correlation, we can conclude that this new methodology successfully determines the vigor and the vegetative expression of vines. With this new methodology it is unnecessary determinate the pruning mass and the number of sticks per plant in the vineyard. It's a simple use application that determines vigor and vegetative expression, based on a digital image of the vine and independent of the pruning date. This new method has a patent register in Portugal, with the reference number 105163.
Related Topics
Physical Sciences and Engineering
Computer Science
Computer Science Applications
Authors
Paula Cristina Oliveira, João Paulo Moura, LuÃs Filipe Fernandes, Elza Maria Amaral, Ana Alexandra Oliveira,