Handle fibers that are very close for 2d PSF analysis

Two fibers on the left-hand side of the detector (2nd and 3rd from the left) are too close for me cleanly separate at large amounts of defocus. My naive algorithm fails at it can not separate these nearby images and I am not currently able to analyze these spots. This ticket is resolve this issue by either:
1. Managing to separate the spots even at large amounts of defocus
2. Managing to separate at medium amounts of defocus and formalizing how to ignore images with this sort of problems