Radial distribution errors over occasional cells

[azagajewski]

I've noticed the radial distribution is occasionally not calculated correctly, especially when the cell is approaching vertical orientation. This is most evident when trying to reconstruct the image. Tracked down to a single bug, have a look at the PR.

To reproduce, initialise over the bitmap and parameters provided in #119 and reconstruct over the binary channel. (The issue is the same with fluorescence channels, not shown here for sake of brevity) - this is what I get, which is not correct.