A system analysis on the studies of UAV images annotations which is releated to semantics

In this study of Thesis, it is aimed that to make a system analysis on the studies of UAV images annotations which is releated to semantics. The thesis consists of; Remote Sensing, UAV, Classification Techniques, The Enrichment of The Fusion of Hyper/Multi-Spectral Images, Object Tracking From UAV, Annotations in UAV images with Geo-referencing.Nowadays, except USA and Israel, UAV producer countries do not have enough information and infrastructure about the robust object tracking based on learning. Firstly, ?Sensor fused night vision?belongs to WAXMAN and ?Fusion of multi sensor imagery for night vision? belongs to FAY is about the multispectral fusion of the images. Secondly, PREDATOR, Tracking-Learning-Detection system which is copyrigted by Google tech., first version is released as open source, is assessed as a robust tracking and detecting system which is a unsupervised learning system from a single frame. Thirdly, ?UAVs in Research? belongs to GOODRICH, from Brigham Young Üniversity and ?Annotations in UAV images with Geo-referencing? belongs to SOLBRIG, from Research Establishment for Applied Science, and ?Automatic photo to terrain alignment for the annotations of mountains? belongs to BABAUD and friends workings are releated to supervised learning in UAV images, in which labels & annotations given to objects with geo reference. İf this model, the transfer of information from data base to UAV images, is modified in accordance with the Turkish Land Forces requirement, it is assessed that it will be much more useful for UAV Operatör, İmage Analyst and Decision Makers which gives them situational awareness.