EVALUATION OF INTERPOLATION METHODS OF AIRBORNE LASER SCANNING DATA FOR DETECTION OF TREES AND THEIR HEIGHTS

This article focuses on the potential of airborne laser scanning data (ALS) in forestry for measuring tree heights and subsequent determination of the total number of individual trees in forest stands. We evaluated different methods of ALS data interpolation acquired with Leica ALS50 -II scanner to a digital surface model (DSM) form. Evaluation was carried out on the basis of the accuracy of elevation models and the determination of total number of trees in forest stands. Results are compared with position and height of the trees acquired by terrain measurement on research plots of 1.25 ha. On these plots the position of all trees and their heights were surveyed using the total station. Firstly, the altitudinal accuracy of ALS for the creation of DSM was evaluated, derived from comparison with terrain measurement. Subsequently, tree tops were detected using hydrological modeling tools in ESRI ArcGIS software and compared with total number of trees on the plots. In accordance with previously published works, our results show that the ALS data interpolations to DSM always underestimate tree heights; when working with DSM of vegetation it is always necessary to use the smoothing filter. When the tree tops and the associated determination of the number of trees are detected in the forest stands it is crucial to choose suitable interpolation and cell size of the resulting grid.