Modeling and removing depth variant blur in 3D fluorescence microscopy

Like many other imaging techniques, 3D fluorescence microscopy suffers from degradations that are basically varying with the depth of the point source. This is due to the light refraction phenomenon. In this article, we focus on modeling and removing depth variant blur in such a system. In particular, we study some of the existing space-variant blur approximations and consider an efficient approximation where the space variant blur function is a linear combination of a set of space-invariant ones. We then focus on restoring space-variant blurred images using such a model. For that, we fit a domain decomposition-based minimization approach to the deconvolution problem with a space variant blur model. We thus obtain a fast restoration algorithm where the image estimation is performed in a parallel way on different sub-images.