End working on the new upsample method, no conclusive results, moving to search/reconstruct divergence

Implementing the new interpolation similar to matlab, working partially, but there is a size problem

Comparaison of files right after peel init, and trying to feed C++ with Matlab data, not working for now

 Removing compilation files, testing for search iterations, multiples debugs and modifying (without consequences) the borders of occ mask during search and reconstruct