Cache-optimized implementation of the filtered backprojection algorithm on a digital signal processor
- 1 October 2007
- journal article
- Published by SPIE-Intl Soc Optical Eng in Journal of Electronic Imaging
- Vol. 16 (4) , 043010-043010-13
Abstract
The reconstruction of an image from its projections through filtered backprojection is not new and has gained popularity since its application on computerized tomography (CT) for medical imaging. However, the number of projections to reconstruct a single image can be in the thousands. Thus, significant computational power is required. Today’s digital signal processors (DSPs) offer an alternative for implementations of a backprojector to process large amounts of data. We demonstrate an implementation in a fixed-point DSP. This implementation is optimized for cache memory to reduce CPU stalls, and we achieved up to 99.6% of effective CPU cycles. With a DSP running at , an image of pixels can be reconstructed from 180 projections in .
Keywords
This publication has 26 references indexed in Scilit:
- Hyperfast Parallel--Beam Backprojection2006
- Introduction to the Cell multiprocessorIBM Journal of Research and Development, 2005
- Computer vision hardware using the Radon transform2002
- A VLSI architecture for two-dimensional Radon transform computations2002
- An image processing IC for backprojection and spatial histogramming in a pipelined arrayIEEE Journal of Solid-State Circuits, 1993
- High-speed computation of the Radon transform and backprojection using an expandable multiprocessor architectureIEEE Transactions on Circuits and Systems for Video Technology, 1992
- Practical cone-beam algorithmJournal of the Optical Society of America A, 1984
- Image Reconstruction from Projections1984
- Introduction1982
- The Fourier reconstruction of a head sectionIEEE Transactions on Nuclear Science, 1974