Deformation of matching imaging

[RainerHeintzmann]

Thanks a lot for this great package on deforming images. I would like to use it for registering 4 matching images for polarization microscopy. However the reference points (e.g. in the first image) are not necessarily on a regular (rectangular) grid, but ideally I would like to use at least a deformed grid (e.g. from landmarks near the corners) or even better any landmark positions. Another application could be multi-color single molecule microscopy, where the chromatic aberrations are important and each filter deforms the image slightly differently but for calibration one could co-register beads at random positions. Therefore two questions:

• Can the current package be used for a deformed grid at the source as well as on the destination side and how to use the interface, as seems is difficult to grasp in the documentation.
• I saw that the idea of a VoroiDeformation is already in the comments of the source. This looks exactly like what I was looking for, at least for the case of reference bead positions. Any plans to flesh this out?

[timholy]

Sorry I didn't notice this. I don't have immediate plans to extend this but I'd be happy to take pull requests. Currently there is no support for non-rectangular grids.