1. Regularization of inverse problems with translation invariant frames.
- Author
-
Göppel, Simon, Frikel, Jürgen, and Haltmeier, Markus
- Subjects
- *
SINGULAR value decomposition , *LINEAR operators , *MATHEMATICAL regularization - Abstract
In various fields of applications, inverse problems are characterized by their sensitivity to data perturbations which can result in serious reconstructions errors. Consequently, regularization techniques have to be employed in order to ensure both stability and reconstruction quality. In recent years, novel approaches have emerged that are able to overcome the limitations of classical methods like filtered singular value decomposition (SVD). One noteworthy development is the utilization of frame-based diagonalization strategies, e.g., given by the wavelet-vaguelette (WVD) decomposition. While these methods can be adapted to specific problem instances, it is well-known that the lack of translation invariance in multiscale systems can introduce certain arti-facts into the reconstructed objects. To address these limitations, we employ the translation-invariant diagonal frame decomposition (TI-DFD) for linear operators. To provide a practical example, we construct a translation-invariant TI-WVD for a one-dimensional integration operator and validate our theoretical insights through numerical simulations. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF