Aircraft-borne Lidar has been used for the accurate measurement of 3D canopy structure and tree height. However, the use of Lidar is costly and is not practical for high-frequency monitoring. Unmanned aerial vehicle (UAV) is a reasonable and convenient system for remote sensing applications. In this study, we examined an effective method to reconstruct digital surface model (DSM) of larch canopy and digital terrain model (DTM) of the underlying ground surface in larch canopy from color images measured by a UAV-borne camera with different focal length lenses (28 mm, 35 mm or 50 mm). Structure from motion (SfM) and surface reconstruction methods such as inverse distance weighting (IDW) and polygon methods were used for the DSM and DTM modeling. In addition, the digital canopy height model (DCHM) was generated by subtracting DTM from DSM. It was found that the tree canopy height estimate using the IDW method and the 28 mm lens was the best with a root mean squared error (RMSE) of 0.47 m.
