This work scrutinises the deconfinement problem of photoswitching microscopy. The task of deconfinement is locating the sources of fluorophoric emission with subpixel accuracy in large sequences of images, which are acquired by temporally confining photoswitching microscopy.

The preceding work on this task is summarised and formulated with a standardised terminology. Then, the novel approach of motivational fitting is introduced, which greatly reduces the number of nonlinear fit attempts that are needed. A complete algorithmic concept and an implementation are given for this approach, along with a visualisation based on weighted histogram equalisation.

The resulting implementation in C++ code is verified and used for extensive testing of the chosen algorithms and of the important alternatives. Major parameters are determined experimentally on both real and simulated data. Additionally, it is proven that the deconfinement problem can be solved well within the real-time domain for a wide range of acquisition speeds and parameter choices. This includes the published parameters for the major temporally confining photoswitching microscopy methods.

This real-time capability alleviates the computational burden of temporally confining photoswitching microscopy and can considerably aid wide-spread use of these methods.