 This study demonstrated that it is possible to use open-source computer vision algorithms to generate high spatial resolution 3D point cloud data sets with RGB spectral attributes from digital photographs taken from a kite aerial platform. The results show that this method can accurately predict tree canopy heights when compared to LIDAR data, although LIDAR has greater precision due to its ability to observe terrain under closed canopy forest. This method could provide a cost-effective way to measure vegetation structure and spectral characteristics at high spatial resolution without the need for expensive equipment or specialized personnel. This article was authored by Jonathan P. Dandois and Earl C. Ellis.