Tree-stem Volume Calculation from SfM-MVS to Improve Estimates in the Timber Industry in Indonesia

Abstract

Spatial Analysis is often thought through the cartographic two dimensions of space, with eventually a 3D component and any form of data associated to it. This concept inherited from eras from data mapping on paper has been driving recent computational development in spatial analysis, however spatial analysis can free itself from this virtual plane, we human evolve on and use any space vertical, reversed, spherical, tubular to investigates all sorts of issues, like the production of timber from trees for instance. As Indonesia is a forestry nation, being able to effectively assess the usable amount of timber available. In the present contribution we aimed to show the loss that can be miscalculated using DBH traditional measurement, and to do so we used a photographic based Structure from Motion method combined with Wavelet Decomposition Analysis. The data shows that the DBH method is correct between 60 to 80%, when considering large-scale variations between DBH and SfM-obtained data, showing that the estimates for the timber industry could be improved, even without the use of expensive laser-based equipment.