Smartphone vs Terrestrial Laser Scanning for hazelnut plant 3D characterization

The geometrical characterization of trees plays an important role in reducing production inputs and improving the sustainability of hazelnut orchards. Recently, digital technologies such as Unmanned Aerial Vehicle (UAV) and Terrestrial Laser Scanning (TLS) have been successfully applied to offer standardized procedures for measuring geometrical parameters, such as canopy height, width, volume, in hazelnut or olive orchard. However, these technologies are expensive and require specific skills, therefore in a hazelnut orchard, a smartphone camera was used to test and validate a procedure for assessing tree geometrical characteristics. Differences between point clouds obtained from TLS and the smartphone were evaluated in two periods: October 2023 (with leaves) and February 2024 (leafless). For the first survey, comparisons were performed on geometrical parameters, including canopy height, tree height, canopy area, and canopy volume, while for the second survey, point-to-point distance comparisons were conducted. The analysis showed a good performance of the Apple iPhone to reconstruct the 3D model of the trees with a coefficient of determination R2 higher than 0.93 for all the measurable variables. Point cloud comparison after the winter dormancy period revealed differences within a few centimeters primarily attributable to sensor and environmental factors. These findings indicate that both technologies can provide accurate tree models, with minor discrepancies generally within acceptable limits for precision agriculture and land management applications.