HAH - got it! Nice overview of what needs to be done.
This time, due to experiences of the past, it's time to make it an Open Source project (GPL) that noone will be able to rip me off.
It'd be based on processing images with a possibility to extend it to live streams - that's better modularity. Core scientific functionality in the core GPL program, then possibly commercial modules to connect with various hardware.
For programming toolkit, I'm considering Kylix (easier) or KDevelop (most versatile). Both Qt based, that'd do fine. Hm, Matthias Kalle Dalheimer has written a nice book on Qt programming. Time for bookstore!
Good Book. Settled on KDevelop in the end, and now have some absolutely pre-alpha code. And have the manual scanned and OCR'ed, too, needs relayout and new illustrations. Lots to do...