Accelerated Image Registration using Grapics Hardware Registration of multi-modal image data is a central part for visualization and analysis of tomographic medical image data. Due to the complexity of the model and the hight number of degrees of freedom, non-rigid registration in particular is a computational expensive and demanding task.
Within this project, methods are to be investigated in order to speed up the registration process considerably by exploiting features of modern, commodity graphics hardware. The mapping of the core parts of the registration algorithm, transformation, similarity measure and optimization, onto graphics hardware should be investigated and an analysis of their efficiency should be performed. | Project manager: Prof. Dr.-Ing. Marc Stamminger, PD Dr. Peter Hastreiter, Prof. Dr. Günther Greiner
Project participants: Prof. Dr.-Ing. Matthias Teßmann
Keywords: Registration;GPU;Graphics Hardware
Duration: 1.9.2007 - 31.10.2010
Mitwirkende Institutionen: Neurozentrum des Kopfklinikums der Universität Erlangen-Nürnberg
Contact: Teßmann, Matthias E-Mail: matthias.tessmann@th-nuernberg.de
| Publications |
---|
Teßmann, Matthias ; Eisenacher, Christian ; Enders, Frank ; Stamminger, Marc ; : GPU Accelerated Normalized Mutual Information and B-Spline Transformation. In: Botha, C.P. ; Kindlmann, G. ; Niessen, W.J. ; Preim, B. (Ed.) : Proceedings of the Eurographics Workshop on Visual Computing for Biomedicine (EG VCBM) 2008 (Eurographics Workshop on Visual Computing for Biomedicine (EG VCBM) 2008 Delft, The Netherlands). Germany : The Eurographics Association, 2008, pp 117-124. - ISBN 978-3-905674-13-2 |
Institution: Chair of Computer Science 9 (Computer Graphics)
|
|