Spatial modeling of coffee tree height and diameter using LiDAR SLAM, RGB imagery, and field measurements
Laen...
Kuupäev
2026
Kättesaadavus
Ajakirja pealkiri
Ajakirja ISSN
Köite pealkiri
Kirjastaja
Estonian University of Life Sciences
Abstrakt
Remote sensing has become established as a promising tool for crop management. In coffee farming, estimating canopy height and diameter is essential to characterize plant vigor, development, and uniformity. This study compared three approaches to obtain these variables: field measurements, RGB (Red, Green, Blue) imagery acquired by a Remotely Piloted Aircraft (RPA), and a handheld Light Detection and Ranging sensor (LiDAR), using geostatistics to assess spatial accuracy and agreement among methods. From three-dimensional data, surface and terrain models were derived, enabling estimation of canopy height and diameter, while geostatistics was used to characterize the spatial structure, select variogram models via cross-validation, and generate continuous maps through ordinary kriging. In pointwise comparison with field data, LiDAR showed better performance for height (R² = 0.80) than the RGB orthomosaic (R² = 0.63). For canopy diameter, LiDAR was also superior (R² = 0.79) relative to RGB imagery (R² = 0.71). At the spatial level, the comparison between kriged maps indicated greater agreement between LiDAR and field data, especially for plant height (r = 0.99), whereas RGB and field data showed lower spatial correspondence (r = 0.44). For canopy diameter, the maps derived from LiDAR and RGB showed good agreement with field data, with r = 0.88 and r = 0.83, respectively. Thus, the combination of pointwise assessment and geostatistical analysis proved suitable for comparing the studied methods and for interpreting, in an integrated manner, the estimation and spatial distribution of coffee tree height and canopy diameter.
Kirjeldus
Received: January 30th, 2026 ; Accepted: July 28th, 2026 ; Published: August 19th, 2026 ; Correspondence: gabriel.ferraz@ufla.br
Märksõnad
digital and precision agriculture, morphometric modeling, point cloud, spatial interpolation, precision phenotyping, articles
Viide
Rubio, S. V., Ferraz, G. A. S., Cardozo, E. J. S., Zavala, E. H., de Oliveira, F. M., Bambi, G., Sarri, D., Reis, G. M., & Ferraz, P. F. P. (2026). Spatial modeling of coffee tree height and diameter using LiDAR SLAM, RGB imagery, and field measurements. Estonian University of Life Sciences. https://doi.org/10.15159/AR.26.060