Innovation in olive-growing by Proximal sensing LiDAR for tree volume estimation

Abstract

The development of methods and techniques for reconstructing olive tree canopies by point clouds is a long-established topic challenging many researchers. Using Mobile Laser Scanner, we measured single tree parameters in an olive grove in Cartoceto, Italy. The area is a well-known geographical indication agreed by Italian law by Protected Designation of Origin (PDO). According to the agronomic practice, we first estimated the canopy volume using geometrical primitives (paraboloid and toroidal) as ground truth. We have finally compared the canopy values with the volumes obtained by mesh algorithms: Convex hull and Alpha shape, to statistically compare pairwise correlation (Paraboloid - Convex hull $\boldsymbol{\mathrm{R}^2=0,92}$ and Toroid - Alpha shape $\boldsymbol{\mathrm{R}^2=0,91)}$. This preliminary analysis provides a theoretical benchmark for future implementations, such as the possibility of including LiDAR (Light Detection and Ranging) in the mechanized pruning process.