Individual tree crown estimation using hyperspectral image and LiDAR data

Detailed tree attributes such as tree height, tree type, diameter at breast height, number of trees are critical for effective management and analysis of the forest. By usage of airborne LiDAR (LIght Detection And Ranging) and hyperspectral data, this paper presents individual tree extraction method to get the accurate tree information. SVM (Support Vector Machine) classifier was used in hyperspectral data classification for extraction of tree area. Then, we performed PCA (Principal Components Analysis) on the hyperspectral image then segmented the test area with LiDAR nDSM (Normalized Digital Surface Model). The results showed that the fusion data provides better results.