| Full text | |
| Author(s): |
Total Authors: 2
|
| Affiliation: | [1] Univ Estadual Campinas, Dept Appl Math, BR-13083859 Campinas, SP - Brazil
Total Affiliations: 1
|
| Document type: | Journal article |
| Source: | Journal of Inverse and Ill-Posed Problems; v. 21, n. 5, p. 695-712, OCT 2013. |
| Web of Science Citations: | 4 |
| Abstract | |
We are concerned with an iterative inversion of the Generalized Attenuated Radon Transform (GART) first introduced by Kunyansky (1992). It is essentially based on approximating the inverse of GART by a scaled inverse Radon transform followed by an appropriate iterative refinement. This paper presents a convergence proof for this iterative method introducing tighter and computable estimates (compared to Kunyansky) for the iteration contraction constants. As expected, such constants depend on the attenuation parameter mu, i.e., the higher mu is, the closer the constant is to one yielding slow convergence to the solution. Such values are compared with those obtained using the infinite-dimensional version of the power method by Krein and Rutman (1950). (AU) | |
| FAPESP's process: | 09/15844-4 - On the inversion of the generalized Radon Transform: new results in fluorescence tomography |
| Grantee: | Eduardo Xavier Miqueles |
| Support Opportunities: | Scholarships in Brazil - Post-Doctoral |