Abstract
LIDAR point clouds from terrestrial or drone-based laser scanners are progressively used in forest inventory and environmental monitoring. Extraction of basic tree features like tree diameter and position can be considered state of the art. The extraction of such features is already implemented in different software tools, these tools does not consider image information during feature extraction.
The extraction of more complex features such as branch-free height, stem quality and 3D shape, stem deformation and damages, vertical and horizontal dead woods, thickness of side-branches, juvenile or herbaceous coverage are still in an experimental stage. Such projects can show a principle feasibility, but software tools for operational application are not yet available.
Umweltdata is holding a patent on an apparatus and method of terrestrial laser-scanning, which avoids occlusion between trees by positioning the scanning sensor on an eccentric protruding slowly rotating arm. The technical feasibility of this scanning method is simultaneously evaluated in a feasibility study. Datasets from simulations of the new method are already available.
The aim of this proposal is to improve existing and developing new feature extraction algorithms by combining the laser detected point clouds and high-resolution images, also providing support for the patented terrestrial laser-scanner. Additional goal is assembling a prototype software from these algorithms and testing it on various datasets collected from wide range of forestry sample plots.
The commercialization of resulting software can be achieved either by packaging with the hardware under development, selling individually or providing cloud processing services.
Hungary