Applications for point cloud skeletonizations in forestry and agriculture

In recent years, both airborne and terrestrial laser scanning developed to a standard technique for acquiring information on terrestrial landscapes. Here forest inventory and orchard management is considered. Inventory and parameterization of individual trees in orchards and forests are of large interest in nowadays society because of the potential of economical maximization of orchard production and the sustainable management of forests. Complex objects like trees require a detailed structural analysis before an analysis of single trees is possible. Skeletonization is such a structural description that enables the extraction of length, diameters, volumes and position of individual branches but also of trees as a whole. In this paper an overview of current possibilities for skeletonization algorithms on trees is given. A new method of skeletonization, especially designed for the analysis of tree data is described and its applicability demonstrated on airborne and terrestrial laser scanning scenarios. The evidence reviewed in this paper leads to conclude that skeletonization is a valuable tool for forest inventory and orchard management. It will be shown that skeletonization offers the possibility of species independent measurement of trees, which make skeletonization a general tool for parameter extraction on trees that empowers also research fields different from orchard management and forest inventory.