Iterative Image Reconstruction From Projections Based On Generalised Kaiser-Bessel Window Functions

Filip Jacobs and Ignace Lemahieu

University of Ghent Sint-Pietersnieuwstraat 41

B-9000 Ghent, Belgium jacobs@petultra.rug.ac.be

Abstract - Tomographic images are calculated from data provided by a scanner. The reconstruction algorithms used for this purpose are either analytical or iterative in nature. The paper focuses on an iterative algorithm called the row-action maximum-likelihood algorithm, and on transmission tomography. Iterative algorithms approximate the image as a linear combination of a limited set of basis functions. The paper introduces a new set of basis functions, called blobs, into the field of process tomography. It also presents preliminary results of a study which evaluates the advantage of using blobs, instead of pixels, for different amounts of available data and different noise levels. The results clearly show that the use of blobs is also beneficial for process tomography.

Keywords : tomography, reconstruction, projection, blobs