Evaluation of a Low-Cost Photogrammetric System for the Retrieval of 3d Tree Architecture

Abstract

Abstract. Reconstruction of major branches of a tree is an important first step for the monitoring of tree sway and assessment of structural stability. Photogrammetric systems can provide a low-cost alternative for the acquisition of three-dimensional data, while also enabling long-term monitoring of a tree of interest. This study introduces a low-cost photogrammetric system based on two Raspberry Pi cameras, which is used to reconstruct the tree architecture for the purpose of stability monitoring. Images of five trees are taken at a range of distances and the resulting point clouds are evaluated in terms of point density and distribution with the reference to TLS. While the photogrammetric point clouds are sparse, it was found that they are capable of reconstructing the tree trunk and lower-order branches, which are most relevant for sway monitoring and tree stability assessment. The most optimal distance for the reconstruction of the relevant branches was found to be 9–10 m, with a baseline of 120 cm.