Lidar-based 3D tree reconstruction enables the retrieval of detailed tree structure; however, many existing methods are based on high-density discrete return lidar datasets. In this paper, we propose the use of small footprint waveform lidar data to achieve branch-level tree reconstruction for both leaf-off and leaf-on conditions. The DIRSIG simulation environment was used for algorithm validation purposes. Leaf-off data served as reference, and leaf-on reconstruction for a particular tree resulted in an average branch length difference of 0.07 m and an average angular difference of approximately 6 degrees for both tilt and azimuth angles. Compared to in situ methods this approach may be used by an airborne system for accurate estimation of forest biomass, forest inventory, land degradation, etc. in large scale applications. Furthermore, since this approach can also be applied on leaf-on trees, the tree skeleton characterization eventually can be conducted year round and will be less dependent on seasonal changes.
No takes yet. Share an insight, caveat, or question.
Wu et al. (2013) studied this question.