Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

SIFT implementation and optimization for general-purpose GPU

SIFT-Implementierung und Optimierung für General-Purpose-Berechnungen mit GPUs
: Heymann, S.; Müller, K.; Smolic, A.; Fröhlich, B.; Wiegand, T.

Rossignac, J. ; European Association for Computer Graphics -EUROGRAPHICS-:
WSCG 2007, The 15th International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision 2007. Full papers proceedings II : University of West Bohemia, Plzen, Czech Republic, January 29 - February 1, 2007
Pilsen: University of West Bohemia, 2007
ISBN: 978-80-8694398-5
ISBN: 978-80-8694301-5
International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG) <15, 2007, Plzen, Czech Republic>
Fraunhofer HHI ()
Merkmalextraktionsverfahren; graphische Datenverarbeitung; Computer Graphik; Parallelverarbeitung; Nachführung; Echtzeitbetrieb; Echtzeitsystem; digitale Bildverarbeitung; Parallelalgorithmus; Rendering (Rechnergraphik); Bilderkennung

With the addition of free programmable components to modern graphics hardware, graphics processing units (GPUs) become increasingly interesting for general purpose computations, especially due to utilizing parallel buffer processing. In this paper the authors present methods and techniques that take advantage of modern graphics hardware for real-time tracking and recognition of feature-points. The focus lies on the generation of feature vectors from input images in the various stages. For the generation of feature-vectors the Scale Invariant Feature Transform (SIFT) method is used due to its high stability against rotation, scale and lighting condition changes of the processed images. The authors present results of the various stages for feature vector generation of their GPU implementation and compare it to the CPU version of the SIFT algorithm. The approach works well on Geforce6 series graphics board and above and takes advantage of new hardware features, e.g. dynamic branching and multiple render targets (MRT) in the fragment processor. With the presented methods feature-tracking with real time frame rates can be achieved on the GPU and meanwhile the CPU can be used for other tasks.
Entnommen aus TEMA