Local refinement search range per iteration? (feature request)


The search ranges for shifts and orientations in local refinement appear to be defined as the maximum amount the orientations or offsets can change by overall, over the entire refinement, rather than per iteration.

Would it be possible to add an additional parameter restricting how much they can change in any one iteration?

I have some local refinements where relatively large search ranges are required to get the best results, but the resulting maps are kind of spiky and over refined around the edges. I am wondering if this could be partly because there is too much freedom for particles to alter their orientations early in refinement. Being able to restrict the search range per iteration would be useful to test this.


1 Like

Hi Oli,

Thanks for requesting this – we have allowing “per-iteration” alignment constraints tracked (i.e. global pose/shift re-initialization at each iteration), and hope to have it available as a feature in a later release.