Fusion of full waveform Lidar and hyperspectral remote sensing data for the characterization of forest stands

摘要

Full waveform small footprint laser scanning und airborne hyperspectral image data from the sensor HyMap of two forest areas in Germany were fused for a detailed forest stand characterization. Fusing active laser scanning data with passive hyperspectral data increases the information content without adding much redundancy. Hyperspectral data offer the maximum spectral reflectance information available from remote sensing. Full waveform laser scanning records the vertical distribution of reflected laser energy and offers the maximum of structural information about forest stands. In order to combine both datasets, we defined voxels above the HyMap pixels, containing the mean laser intensity in slices of 50 cm or 100 cm height for the area of each HyMap pixel. These datasets yield a detailed impression of the vertical structure of forest stands and can be used, among other things, to derive tree height, crown base height, and biomass. In addition, the joined images performed better in classifying tree species and age classes than each of the single images.