Hi! Newbie..couple of questions..

[Physicist13]

Hello All. Currently on the Trial of PI and so far loving it! Def appeals to my scientific nature 
I have a couple of questions that so far I cant see have been addressed but this may be wrong!
1) The noise minimisation via an Optimisation routine: What Optimisation algorythm is used? What about the local/global minimum problem? 2) Scaling of darks is linear, thus only first term of Exp(kT)? how about the next term to give wider T range? 3)  new Align has Geometric correction..script to generate the correction vector CSV? could this in principle be used to correct for Coma??
Many thanks!
Patrick

[Carlos Milovic]

1) I assume that you are talking about TGVDenoise. It uses the Chambolle algorithm, where the minimization problem is expanded with the dual to a concave-convex problem, and a saddle point is searched.


Chambolle, A., Pock, T. A first-order primal-dual algorithm for convex problems with applications to imaging. J. Math. Imaging Vis. 40, 120–145, 2011.

[Physicist13]

Carlos, thank for that info! that useful.  I should have been clear. I was refering to the calibration phase of Dark scaling to reach a minimum residual noise. Was wondering how that particular optimisation worked?
thanks,
Patrick