Assessing models from Lidar based vegetation indicators for predicting spraying deposit amounts in a set of vine estatesin France

摘要

Lidar 2D sensors mounted on a tractor can be used to calculate grapevine crop structure indicators for a reasonable cost. Previous work highlighted the relevance of two indicators based on 2D Lidar for predicting intercepted spray deposits and thus facilitating appropriate dose adjustment. The indicators considered were the Tree Area Index (TAI) and the Leaf Wall Area by Point (LWApts). TAI is based on a stochastic expression of interception. LWApts is based on vegetation geometry and was meant to bea high resolution variant of classic Leaf Wall Area. Models based on experimental data were calculated for two sprayer types, for predicting deposits from TAI and LWApts. In the present study, previously learned models are assessed on a new and larger set of data (other plots, other year). Also, the variability of TAI and LWApts is studied on 20 plots from four French vine estates of different production areas.