Assessing plant area index and 3D plant area distribution in wheat canopies from a terrestrial LiDAR system for high-throughput plant phenotyping
Abstract
This study investigates a new methodology to estimate plant area index (PAI) and its spatial distribution from 3D point clouds acquired by a system of 3 LiDARs (2 nadir-viewing and one at 45°) mounted on the Phenomobile Unmanned Ground Vehicle designed for high-throughput phenotyping of herbaceous crops. It relies on the computation of the gap fraction at the voxel level using the trajectories of the laser beams to numerically invert the Beer-Lambert law (BL), enabling the simultaneous estimation of PAI and AIA (Average Inclination Angle, in°). The method is evaluated on wheat crops both in silico (i.e. simulations based on the AdelWheat functional structural crop model coupled with a point cloud simulator) and in actual phenotyping trials of a panel of 10 bread wheat genotypes grown in two study sites under different treatments of plant density, water stress and sowing dates. The in silico validation indicates that the numeric inversion of BL with the LiDAR system of the Phenomobile leads to a relative error of 9.8% (RMSE = 0.32, R2 = 0.99) and 8.4% (RMSE =5.4°, R2 = 0.95) in the estimation of PAI and AIA, respectively. The validation in actual phenotyping experiments against destructive measurements produced satisfactory results for canopy PAI (RMSE = 1.44, R2=0.82), while requiring prior knowledge on AIA, e.g. either by (i) introducing a regularization term in the cost function used to invert BL, to avoid unexpected high AIA estimations, (ii) using a fixed, reference value of 60°. The analysis of the results indicates that the inclusion of a regularization term mitigates the impact of actual differences in leaf inclination among cultivars on PAI accuracy, while providing a relative description of the actual variability in the vertical profile of leaf inclination across the different cultivars.
Related
- Spall Failure of Coral Skeleton beneath Gas-Laden Canopies: An Idealized Blast-Fishing Model
- Editorial: Dynamic exospheres of terrestrial bodies through the solar system
- Precession and meltwater control of greenhouse gas emissions from terrestrial sources during the last glacial period
- North Atlantic chlorophyll blooms modulate European hydroclimate and terrestrial productivity
- More concentrated precipitation decreases terrestrial water storage
BibTeX
@article{lopezlozano2026assessing,
title = {{Assessing plant area index and 3D plant area distribution in wheat canopies from a terrestrial LiDAR system for high-throughput plant phenotyping}},
author = {Raul Lopez-Lozano and Tian Ma and Maxime Soma and Marie Weiss and Aurelien Ausset and Philippe Burger and Bruno Berthon and Romain Chapuis and Marie-Pia d'Argaignon and Antonin Grau and Florian Larue and Romane Le-Roy and Remy Marandel and Vincent Mercier and Mathieu Roy and Gilles Tison and Frederic Venault and Pierre Martre and Frédéric Baret},
journal = {Remote Sensing of Environment},
year = {2026},
month = oct,
doi = {10.1016/j.rse.2026.115711},
url = {https://hal.inrae.fr/hal-05776009v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Assessing plant area index and 3D plant area distribution in wheat canopies from a terrestrial LiDAR system for high-throughput plant phenotyping
AU - Raul Lopez-Lozano
AU - Tian Ma
AU - Maxime Soma
AU - Marie Weiss
AU - Aurelien Ausset
AU - Philippe Burger
AU - Bruno Berthon
AU - Romain Chapuis
AU - Marie-Pia d'Argaignon
AU - Antonin Grau
AU - Florian Larue
AU - Romane Le-Roy
AU - Remy Marandel
AU - Vincent Mercier
AU - Mathieu Roy
AU - Gilles Tison
AU - Frederic Venault
AU - Pierre Martre
AU - Frédéric Baret
PY - 2026
DA - 2026/10/02
JO - Remote Sensing of Environment
DO - 10.1016/j.rse.2026.115711
UR - https://hal.inrae.fr/hal-05776009v1
AB - This study investigates a new methodology to estimate plant area index (PAI) and its spatial distribution from 3D point clouds acquired by a system of 3 LiDARs (2 nadir-viewing and one at 45°) mounted on the Phenomobile Unmanned Ground Vehicle designed for high-throughput phenotyping of herbaceous crops. It relies on the computation of the gap fraction at the voxel level using the trajectories of the laser beams to numerically invert the Beer-Lambert law (BL), enabling the simultaneous estimation of PAI and AIA (Average Inclination Angle, in°). The method is evaluated on wheat crops both in silico (i.e. simulations based on the AdelWheat functional structural crop model coupled with a point cloud simulator) and in actual phenotyping trials of a panel of 10 bread wheat genotypes grown in two study sites under different treatments of plant density, water stress and sowing dates. The in silico validation indicates that the numeric inversion of BL with the LiDAR system of the Phenomobile leads to a relative error of 9.8% (RMSE = 0.32, R2 = 0.99) and 8.4% (RMSE =5.4°, R2 = 0.95) in the estimation of PAI and AIA, respectively. The validation in actual phenotyping experiments against destructive measurements produced satisfactory results for canopy PAI (RMSE = 1.44, R2=0.82), while requiring prior knowledge on AIA, e.g. either by (i) introducing a regularization term in the cost function used to invert BL, to avoid unexpected high AIA estimations, (ii) using a fixed, reference value of 60°. The analysis of the results indicates that the inclusion of a regularization term mitigates the impact of actual differences in leaf inclination among cultivars on PAI accuracy, while providing a relative description of the actual variability in the vertical profile of leaf inclination across the different cultivars.
SN - 0034-4257
SN - 1879-0704
ER -