is there any alternative way ?

I want to remove those part of my data which are higher than 0.75 and those that lower than 0.30 I am using the following way, do you have any other idea?

[X Y Z] = size(H_cube);
  avg_spec = zeros(X,Y);
  specular_mask = zeros(X,Y);
  data_mask = zeros(X,Y);
  counter = 1;
  for i=1:X
      for j=1:Y
          avg_spec(i,j) = mean(H_cube(i,j,:));
          for k=1:Z
              if(H_cube(i,j,k)>0.75 || avg_spec(i,j) >0.30)
                  specular_mask(i,j) = 1;
                 end
              if(H_cube(i,j,k)>0.40)
                  data_mask(i,j) = 1;
              end
          end
       end
  end

Answer 1

[Sean de Wolski]

Do the whole avg_spec calculation at once:

avg_spec = mean(H_cube,3); % mean along third dimension

Use logical indexing to create data_mask and specular_mask

specular_mask = bsxfun(@or,H_cube>0.75, avg_spec>0.30);  % apply or with avg_spec to every page of H_cube

data_mask = H_cube > 0.40; % logical indexing directly