Open esrivas17 opened 2 months ago
Hello, I think the reason for doing this is because in the real world, adjacent pixels have more similarity in physical representation, and in SAR images, because the resolution is mostly a few meters or more, non adjacent pixels may have sudden changes in physical representation. This is my own understanding. If there is anything inappropriate, please forgive me
Hi, I am trying to understand the SHPs calculation implementation. I have a simple question about
SHP_SelPoint.m
It seems that you take the SHPs from a pixel only if these are connected to the pixel coordinate within the search window. (lines 64-65)% connection component
LL = bwlabel(SeedPoint);
SHP.PixelInd(:,num)=LL(:)==LL(InitRow,InitCol);
I wonder, why don't take all the pixels within the search window who pass the statistical test as SHPs, regardless if these are not connected spatially to the pixel?
Thanks a lot for your answer, Erik