clarification of NCC and ZNCC

[johnnychen94]

Originally posted by @mateuszbaran in #44

By the way, your definition of NCC seems more related to ZNCC? See here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3596837/ . Implementation and the docstring are inconsistent about subtracting the mean.

Our NCC implementation is actually zero-mean NCC.

A safe patch to correct the definition is:

• [x] deprecate NCC/ncc in favor of ZNCC/zncc (#65)
• [ ] remove the NCC symbol in ImageDistances v0.3.0
• [ ] add back NCC in ImageDistances v0.4.0 (or maybe even later than that)

[mateuszbaran]

This sounds like a good plan :slightly_smiling_face: .