In this paper four different delineation methods based on airborne laser scanning (ALS) and hyperspectral data are compared over a forest area in the Italian Alps. The comparison was carried out in terms of detected trees, while the ALS based methods are compared also in terms of attributes estimated (e.g. height). From the experimental results emerged that ALS methods outperformed hyperspectral one in terms of tree detection rate in two of three cases. The best results were achieved with a method based on region growing on an ALS image, and by one based on clustering of raw ALS point cloud. Regarding the estimates of the tree attributes all the ALS methods provided good results with very high accuracies when considering only big trees.

This paper presents a 3D delineation method for airborne laser scanning point cloud. The method is based on an unsupervised clustering technique applied on horizontal slices followed by vertical merging based on overlapping among clusters. On an Alpine forest dataset, we analysed the effects of different forest structures and point cloud densities on tree crown delineation. Forest structure affects mainly the omission error, which eases with homogeneous tree spacing and sizes, while on the commission error forest structure has only slight effect. Delineation accuracy increases with higher point densities where MannWhitney-Wilcoxon test shows that accuracy differences between thinned data and original data are statistically significant.

No abstract