Tree identification by laser scanning

LIDAR, which stands for Light Detection and Ranging, is a remote sensing method that uses light in the form of a pulsed laser to measure variable distances. 

LiDAR provides measurements of the horizontal and vertical vegetation structure of ecosystems. The light pulses generate precise, three-dimensional information that, alone, or in combination with satellite multispectral images, allows to automatically and accurately predict forest characteristics, such as tree height, single tree detection, stem diameter, basal area, stem volume, biomass etc.

A joint research project by the Tampere University of Technology's mathematics laboratory and the Natural Resources Institute Finland has developed a new method of identifying tree species based on laser scanning measurements. 

With their method, individual trees can be extracted from forest plot level point cloud data, and the structure of their crowns can be reconstructed as comprehensive 3D models. The created tree models consist of consecutive cylinders, which determine the structure of the tree stem and branches as well as the branching structure.

Previously, it was possible to make a rough distinction between the stem and the crown, based on the point cloud. Now, we are able to make out individual branches and analyse the characteristics of their diameters, volumes and branch angles.

For species identification, the researchers defined 15 classification features, the values of which were then calculated for each tree. Some of these features are completely new and some have been used in previous studies. The new aspect is that now their value can be calculated more accurately, as the colleagues were able to utilize information about the tree's entire crown. They tested three of the most common tree species in Finland, birch, pine and spruce, but they already plan to extend their test set to more species.

According to our results, automatic species recognition is possible with more than 95% accuracy. The purpose was not to find the best possible combination of features, but only to prove that classification based on detailed tree models is possible. However, several combinations produced good results and all the classification methods had a maximum accuracy over 95%. The results also showed that just 30 trees per species is enough learning material for the classification.

Future tests will also include measurements taken from more diverse forests. The tree models calculated based on the laser scanning data will be stored in a database, which can be accessed for even more accurate species recognition when the number of included samples grows.