Question towards the displacement grid positions (m)

[FrankFR]

Dear Sir or Madam,

I read the answer to the questions "Retrieving mesh" coordinates and "How do the coordinates in the displacement matrices, u{1}{1:3}, correlate to the coordinates of the original volume?", but I still don't quite get it.

Imagine I have a reference and deformed volume of size 256x256x256. After running FIDVC (output mesh spacing dm=8 and sSize = 64 ->87,5% overlap of the interrogation windows) the resulting displacement fields u{1,1}{1,1 to 1,4} will have a size of 33x33x33 and like mohakpatel stated:
1 : dm : size(I,1)+dm gives me the positions where the displacement is calculated as m = 1,9,17,25,...,257.

Now, the grid position 257 is the center of the last interrogation window, but outside of my original volume?! Is this correct or am I minsinterpreting the positions where the displacement grid is calculated?

Regards,

Frank

[FrankFR]

I edited the first version of my question due to the fact that I think I understood parts of what I was asking before.

[ALandauer]

Hi Frank, Basically, yes. To reduce errors near the edges, while still getting some information from the DVC, the algorithm pads images and selects subset that may contain padded area. In this case, a subset is centered at m=257; the subset will contain approximately half padding and half image content. For this reason, the information content of edge subsets is lower and we typically see a lot more noise in the resultant displacements.

[FrankFR]

Hello Alex,

thank you for the clarification.

Regards, Frank