Institut für Mathematische Stochastik

SFB755 Nanoscale Photonic Imaging
Subproject - A4 -
Data driven image selection for photonic counting data

Partner:

Project Management:

  • Prof. Dr. A. Munk, Göttingen

Staff:

  • Dr. Jörn Dannemann

Description:

 

Image reconstruction methods based on Poisson data from nano-scale (fluorescence) microscopy are based on regularization techniques to deconvolve the observed data. They require a proper selection of tuning parameters such as the stopping iteration in Richardson-Lucy deconvolution. Its automatic, i.e. purely data driven, choice is a major challenge in image processing. We will develop methods for Poisson data, with stochastic multiresolution analysis as the primary tool.