Automatic stem mapping
A scan line continuousness segmentation method was developed for plot level forest inventories using single-scan terrestrial laser scanning. Two 1D profiles, in horizontal and vertical directions, are constructed and overlaid to segment the data into homogenous segment. For more details on the scan line continuousness segmentation, please see the article Liang et al., 2008. Plot-level trunk detection and reconstruction using one scan-mode terrestrial laser scanning data.
The original intensity image (left) and the clustering result (right).
A robust stem location mapping method was developed for plot level forest inventories. The stem points were identified by point's spatial properties and the 3D stem model. The spatial properties were studied for each point in its neighbourhood. A local coordinate system was set up and the axes directions in the system were used to detect stem points. The possible stem points were grouped. The stem model, a series of 3D right circular cylinders along the stem profile, was built by a robust modelling procedure. For more details on the robust stem location mapping, see Liang, X., Litkey, P., Hyyppä, J., Kaartinen, H., Vastaranta, M. and Holopainen, M., 2012. Automatic stem mapping using single-scan terrestrial laser scanning. IEEE Transactions of Geoscience and Remote Sensing 50(2).
From left to right: original point cloud, detected stem points and the built stem model. On the right an example of a section of stem points and a fitted cylinder model element.