The video shows the complementary project part of a Bachelor-Thesis with focus on extensive research in the areas of depth-cameras and 3D-reconstruction.
The goal of this project was to create a mobile, MAV-based, 3D-scanner to autonomously create 3D-models of observed scenes. While the original focus of the project was to reconstruct full human models, the design of the system aimed for allowing usage in any scene.
Reconstruction of the observed scene was achieved by depth and RGB-image data acquisition during flight. This data was then transferred to a desktop computer, where 3D-model creation was achieved throughout the "ReconstructMe" framework. While there exist similar solutions, involving cheap hardware only, these typically require the user to actively guide the reconstruction. In such cases the camera has to be moved in a specific way, which is not necessary for the proposed solution.
