Averaging method with sliding window

m =100; %vertical window size in samples; n =50; %shift

B=zeros(nz,nx+n,ny); B(:,1:nx,:)=Bnew;

M=zeros(nz,nx+n,ny); M(:,1:nx,:)=Mnew;

B1=reshape(B,[nz ny*(nx+50)]); M1=reshape(M,[nz ny*(nx+50)]);

    fun=@(x) mean(x(:));
    tic;
    scaleB=nlfilter(B1(:,:,:),[m n],fun);
    scaleM=nlfilter(M1(:,:,:),[m n],fun);
    toc;

I am trying to find a scaling factor btw Bnew and Mnew...using a sliding window in nlfilter however, the averaging is too detail. Is there a way to output an averaging scalar across the window for a certain interval. I need the scaling factor to vary slowly.

Thank you !

Answer 1

[Image Analyst]

What does "btw Bnew and Mnew" mean? What is "btw" - between? Between what? Or "by the way"?

Anyway, I don't see any scaling factor sent into fun, so I'm not sure how you would want to use it if you had one. You're not even doing a non-linear filter so there's no reason to use a super flexible function like nlfilter(). Why not just use convn()? That will blur a 3D array very efficiently - convn() is highly optimized.