Articles
Testing the suitability of a terrestrial 2D LiDAR scanner for 3D canopy characterisation of narrow vineyards to optimise the spraying process
Article number
1360_4
Pages
29 – 36
Language
English
Abstract
One strategy to optimise spraying is to adjust the spray rate according to local vegetative characteristics.
Mobile 2D LiDAR proximal sensor data and local measurements of deposition rates from a side-by-side sprayer were made across 11 narrow inter-row fields in three French vineyards at three dates in 2021. Primary canopy attributes (height, width and porosity) and an integrated indicator of canopy geometry (leaf wall area) were calculated from the LiDAR data using a Bayesian point cloud classification algorithm.
Multivariate models to predict the deposition distribution, as deciles, using the primary canopy attributes were constructed and calibrated using data from five of the fields and validated against the data from the six other fields.
The models are based on log-lin regression for data acquired at three growth stages covering the whole season.
The prediction quality and uncertainty of these multivariate statistical models at different growth stages were evaluated by comparison with previously proposed univariate deposition models based on integrated indicators at the same growth stages.
The results showed that multivariate models can predict the distribution of deposits from a typical face-to-face sprayer more accurately (0.79<R2<0.91), and robustly (11.2%<nRMSEp<25.3%) than univariate prediction models based on integrated indicators.
These predictive multivariate models could enable variable-rate sprayers to adjust the spray rate according to local vegetative characteristics in an automated way.
This work is an extension of previous work made on vineyards with wider inter-rows.
Mobile 2D LiDAR proximal sensor data and local measurements of deposition rates from a side-by-side sprayer were made across 11 narrow inter-row fields in three French vineyards at three dates in 2021. Primary canopy attributes (height, width and porosity) and an integrated indicator of canopy geometry (leaf wall area) were calculated from the LiDAR data using a Bayesian point cloud classification algorithm.
Multivariate models to predict the deposition distribution, as deciles, using the primary canopy attributes were constructed and calibrated using data from five of the fields and validated against the data from the six other fields.
The models are based on log-lin regression for data acquired at three growth stages covering the whole season.
The prediction quality and uncertainty of these multivariate statistical models at different growth stages were evaluated by comparison with previously proposed univariate deposition models based on integrated indicators at the same growth stages.
The results showed that multivariate models can predict the distribution of deposits from a typical face-to-face sprayer more accurately (0.79<R2<0.91), and robustly (11.2%<nRMSEp<25.3%) than univariate prediction models based on integrated indicators.
These predictive multivariate models could enable variable-rate sprayers to adjust the spray rate according to local vegetative characteristics in an automated way.
This work is an extension of previous work made on vineyards with wider inter-rows.
Authors
A. Cheraiet, J. Taylor, S. Codis, P. Petitot, F. Bidaut, M. Liebart, O. Naud
Keywords
3D point clouds, variable-rate spraying, precision viticulture
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)