How to divide a 3d matrix by a 2d matrix along one dimension

Hi everybody,

I have 3d Matrix named Data, consisting of several grey scale Pictures. I want to divide it by a 2d Matix called DataMax containing the maximum intensity among the pictures at a pixel, in order to normalize Data in the 3rd dimension to 1.

I can perform the operation using a for loop:

 DataMax = max(Data,[],3);
 [x,y,z] = size(Data);
 for k=1:z
     Data(:,:,k) = Data(:,:,k)./DataMax;
 end

However I would like to get rid of the loop in order to maximize performance.

Any ideas on how to do this?

Answer 1

[Star Strider]

See if bsxfun (link) will do what you want:

Data = bsxfun(@rdivide, Data, DataMax);

It is generally quite efficient.

[Tillmann Spellauge]

Works like a charm! Thanks a lot!

[Star Strider]

As always, my pleasure!