Simulation of airborne and UAV remote sensing data with computer graphics tools

Ground-truth data are an important need for researchers working in the field of remote sensing. Obtaining these data from residential areas, mountainous areas etc. are an expensive and time consuming process. In the scope of this study, 3-dimensional world creation tools and photo-realistic image production tools from this 3D world will be used to obtain images from many different geographical regions and different shooting conditions (such as sunny, slightly sunny, foggy, etc .). Some data will be generated by random and procedural methods, and some data will be sourced from existing open sources. With the blending of all these data, images from synthetically produced regions with very rich, realistic and terrestrial accuracy will be taken with different flight scenarios and will be presented to the service of researchers working in the field of remote research. Thanks to the high data volume that has become possible with this study, it will be more possible to train deep learning methods that need a lot of data and to provide a generalized learning.