 Ecological remote sensing using unmanned aerial vehicles, UAVs, and computer vision structure from motion, SFM. Algorithms has been rapidly adopted, but its accuracy remains uncertain. This study examined the impact of varying lighting, altitude, and image overlap on the accuracy of canopy height estimates derived from UAV-SFM remote sensing. Under optimal conditions of clear lighting and high image overlap, greater than 80%, the study found that the accuracy of canopy height estimates was comparable to those obtained with field measurements and lidar. The study also showed that the variation in point cloud quality was related to the behavior of SFM image features, which may be the fundamental unit of SFM remote sensing. This article was authored by Jonathan P. Dandois, Marco Lano, and Earl C. Ellis.