Voxel extremum filtering and QuickHull: a rapid algorithm for individual tree canopy volume estimation
Year: 2026
Type: article
Abstract: Estimating individual tree canopy volume from dense point clouds is computationally demanding. This study integrates three-dimensional voxel indexing with QuickHull and introduces two filtering strategies: Single-direction Maximum and Minimum Point filtering (SMMP) and Multi-directional Maximum and Minimum Point filtering (MMMP). Both strategies retain local extrema to reduce the point set used for convex hull construction. Tests using terrestrial laser scanning data from three tree species showed that, for point clouds with 1 cm resolution, SMMP at a voxel grid side length of 5 cm reduced computation time to 18.28%–36.68% of that required by standard QuickHull, with volume loss below 1%. At a grid side length of 10 cm, MMMP reduced computation time to 34.55%–73.55% of the standard QuickHull time, with volume loss below 0.5%. SMMP provided greater computational efficiency, whereas MMMP produced lower volume loss. For point cloud resolutions of 10 cm or finer, the recommended voxel grid side lengths were approximately five and ten times the point cloud resolution for SMMP and MMMP, respectively. These methods accelerate convex hull-based canopy volume estimation while maintaining close agreement with standard QuickHull, supporting efficient three-dimensional green quantity assessment and forestry remote sensing.
Authors: Jie Yang, Huan Ding, Bolin Li
Institution: Henan Polytechnic University
Language: English
Subfield: Environmental Engineering
Field: Environmental Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $1,060
OpenAlex ID: W7223241070