This work addresses the geo-localization problem of an Unmanned Aerial Vehicle (UAV), or aerial drone, flying over a densely forested region in the Brazilian Amazon using Computer Vision. LiDAR cloud point data were post-processed to become 2D template windows, and by applying the template matching algorithm against a geo-referenced elevation model from the interferometric synthetic aperture radar (InSAR) satellite data, positions were estimated. The template matching used the Normalized Cross-correlation method. The LiDAR 3D cloud was processed through binning to become 2D 12.5 m pixel spacing surface images, compatible with the ALOS terrain reference map. This application of template matching achieved 48.5 m root mean square error, or less than four pixels, over a simulated route. Position estimation is essential for automatic navigation of aerial vehicles, and experiments with LiDAR data show potential for localization over densely forested regions, where methods using optical camera data may fail to acquire distinguishable features.