State-of-the-art image reconstruction techniques for multi-channel inverse problems employ various inter-channel correlation and regularization techniques to address low data signal-to-noise ratio (SNR) and reconstruction artifacts. This can be imposed through statistical noise dependencies between channels or regularisation penalties employing joint structural regularities across the channel space. For the latter case one can use various vector and matrix norms enabling joint directions of smoothing. This mini-symposium will demonstrate current tendencies in multi-channel image reconstruction through correlative regularisation approaches as well as applications such as spectral computed tomography.

Multi-channel high-resolution x-ray tomography

Doga Gursoy (Argonne National Laboratory)

Collaborative Regularization Models for Color Imaging Problems

Catalina Sbert (Universitat de les Illes Balears)

Electron tomography combining spectral and non-spectral modalities

Zhichao Zhong (Centrum Wiskunde & Informatica)

Joint image reconstruction of multi-channel X-ray computed tomography data for material science

Daniil Kazantsev (University of Manchester)

Organizers:

Jakob Jorgensen (University of Manchester)

Daniil Kazantsev (University of Manchester)

Keywords:

computed tomography, image reconstruction, inverse problems, nonlinear optimization