Articles
Use of VIS-NIR data for characterizing grapevine canopies: remote and proximal sensing approaches at individual vine scale
Article number
1360_41
Pages
339 – 346
Language
English
Abstract
Precision viticulture (PV) relies on the estimation of inter- and within-field variability to define homogenous zones within the vineyard.
Zoning in vineyards can be conducted through the measurement and geo-referencing of canopy parameters, such as leaf area, vine vigour, canopy size and pruning weight.
Grapevine canopies can be characterized by manual measurements, but this method is labour intensive and time-consuming.
The use of remote and proximal sensing methods for PV has been significantly increased in the last years.
With the current technology, it is possible to acquire high spatio-temporal resolution and georeferenced imagery from different sensors and platforms.
In this context, the identification of individual vines is important to precisely assess the vineyard conditions by estimating several metrics for each vine, considering the effect of missing or infected plants.
In this study, a comprehensive field experiment was carried out to test the ability of the VIS-NIR spectral information acquired close to the canopy in a lateral position (hand-held camera proximal approach) and from an unmanned aerial vehicle (UAV) (remote approach) to identify vines with different canopy sizes in a commercial Cabernet Sauvignon vineyard block trained on a vertical shoot-positioning (VSP) system located in Stellenbosch South Africa.
The suitability of the methods was tested using ground truth data (destructive leaf area and pruning weight) and digital image analysis to determine the accuracy and spatial agreement of the estimations.
This type of analysis is critical for PV applications since valid and reliable input data are required to provide proper management decisions.
Zoning in vineyards can be conducted through the measurement and geo-referencing of canopy parameters, such as leaf area, vine vigour, canopy size and pruning weight.
Grapevine canopies can be characterized by manual measurements, but this method is labour intensive and time-consuming.
The use of remote and proximal sensing methods for PV has been significantly increased in the last years.
With the current technology, it is possible to acquire high spatio-temporal resolution and georeferenced imagery from different sensors and platforms.
In this context, the identification of individual vines is important to precisely assess the vineyard conditions by estimating several metrics for each vine, considering the effect of missing or infected plants.
In this study, a comprehensive field experiment was carried out to test the ability of the VIS-NIR spectral information acquired close to the canopy in a lateral position (hand-held camera proximal approach) and from an unmanned aerial vehicle (UAV) (remote approach) to identify vines with different canopy sizes in a commercial Cabernet Sauvignon vineyard block trained on a vertical shoot-positioning (VSP) system located in Stellenbosch South Africa.
The suitability of the methods was tested using ground truth data (destructive leaf area and pruning weight) and digital image analysis to determine the accuracy and spatial agreement of the estimations.
This type of analysis is critical for PV applications since valid and reliable input data are required to provide proper management decisions.
Authors
C. Poblete-Echeverría, A. Berry, J. Luus, M.A. Vivier
Keywords
multispectral imagery, vine detection, spatial variability, unmanned aerial vehicle
Groups involved
- Division Precision Horticulture and Engineering
- Division Plant-Environment Interactions in Field Systems
- Division Vine and Berry Fruits
- Division Tropical and Subtropical Fruit and Nuts
- Division Plant Genetic Resources, Breeding and Biotechnology
- Division Temperate Tree Nuts
- Division Temperate Tree Fruits
- Division Vegetables, Roots and Tubers
- Division Sustaining Horticulture in a Changing World
- Working Group Mechanization, Digitization, Sensing and Robotics
Online Articles (45)