To see the forest for the trees; tree identifying algorithms for Airborne and Terrestrial LIDAR

Organised by Laboratory of Geo-information Science and Remote Sensing

Thu 26 June 2014 11:00 to 11:30

Venue Gaia, gebouwnummer 101

By Joost van der Gun (the Netherlands)

Individually located trees can provide useful information for monitoring deforestation and forest degradation. This thesis aimed to research the detection of individual trees from both Terrestrial and Airborne laser scanning data. Four different tree detection methods were executed. Three of the tree detection methods were based on tree top detection, peak tree top detection, neighbourhood maximum and height dependent treetop detection. One tree detection method was new and based on stem detection, which is called hollow tree stem detection.