Tomographic Reconstruction with Simultaneous Error Correction
Abstract: The advanced imaging technique, which leads to unquestionably improved clarity of the sample structure, is, however, highly vulnerable to systematic and random errors. These errors are fundamental obstacles to a full realization of next-generation photon science since they can lead to reduced spatial resolution and even misinterpretation of the underlying sample structures.
In this work, we focus on the most common experimental error, the center of rotation shift, and we formulate new physical models to capture experimental setups. Then we devise new mathematical optimization formulations for robust inversion of complex samples. The numerical results are demonstrated on both synthetic and real data.