C2 Sparse/Synthetic THz-Imaging for 3D-Tomography

Synthetic image reconstruction methods can solve central problems of THz-imaging, in particular the necessity for large lenses for imaging from large distances, the low depth of imaging systems, and the problems of tomographic analyses of internal structures under consideration of deflection. These basic problems, which direct imaging systems cannot cope with, can be solved by means of synthetic and/or tomographic imaging approaches. This is done at the expense of  considerable additional numeric and technical effort. Because of the extreme numerical expense conventional systems are not able to realize 3D reconstructions in real time. Methods developed in other frequency ranges, as e.g. computertomographic reconstruction, are not directly transferable to the THz-frequency range, due to deflection limitations and diffusion. Thus, a simple adaption of existing methods is not sufficient and new technological approaches have to be found, from system development to the design of new reconstruction methods. Within the framework of the first project phase two systems for full 3D synthetic reconstruction could be realized, one rasterized system at 600 GHz and one system real-time construction at 100 GHz with a hybrid approach, i.e. with one imaging as well as one synthetic dimension. In this context first approaches for the modeling of diffusion on rough surfaces were developed.