In this seminar I will give an overview of some recent results and ongoing projects within limited-data X-ray tomographic reconstruction and sparsity regularization.
First, I will describe a special limited angle tomography problem in which metal bar sample holders cause cone-shaped parts missing in the sinogram. I will describe our attempts of characterizing and preliminary results of removing the resulting artifacts.
Next, I will present a projection matching algorithm for automatically aligning tomographic projections of unknown exact orientation. In the present case we are interested in ultra-high resolution tomographic reconstruction from ptychography data. Currently, unknown sample translations and rotations at the nano-scale for example caused by vibration limit the achievable reconstruction quality. I will show preliminary results with simulated and real synchrotron ptychography data.
Finally, I will mention our attempts of establishing a connection between image sparsity and the amount of undersampling admitted by sparsity-regularized reconstrution methods. In particular, I will describe simulation results suggesting how one may predict the sufficient number of projections to acquire based on the sparsity level of the scanned object. I will then point to a potential follow-up study for validation on real data using the SophiaBeads data set recently released from the University of Manchester.