Prev: fzero logical error
Next: arithmetic on ROI in scatter plots
From: Neha Kaul on 11 May 2010 17:28
Hi,

I'm an inexperienced matlab user and I'm trying to find the standard deviation for each pixel in a sequence of images. Each image is 512x512 and between 500-1000 frames long, so something like this is computationally expensive...
mat1 = zeros(512,512,100,'uint16');
for readin = 1:100
[mat1(:,:,readin)] = imread('002.tif',readin);
end
mat1std = std(double(mat1),1,100);

I would really appreciate any help or ideas for me to pursue.

Thanks!
Neha
From: ImageAnalyst on 11 May 2010 17:36
Neha
Check this out from just yesterday:

http://www.mathworks.com/matlabcentral/newsreader/view_thread/281746#743890

From: Walter Roberson on 11 May 2010 17:54
Neha Kaul wrote:
> Hi,
>
> I'm an inexperienced matlab user and I'm trying to find the standard
> deviation for each pixel in a sequence of images. Each image is 512x512
> and between 500-1000 frames long, so something like this is
> computationally expensive...
> mat1 = zeros(512,512,100,'uint16');
> for readin = 1:100
> [mat1(:,:,readin)] = imread('002.tif',readin);
> end
> mat1std = std(double(mat1),1,100);
>
> I would really appreciate any help or ideas for me to pursue.

In that std() call, the 100 should be 3, as you wish to take the standard
deviation along the third coordinate, not along the 100'th coordinate. You
will not receive a run-time error in asking for the 100'th coordinate, as
every array has an implicit "x 1" indefinitely often. Your 512 x 512 x 100
matrix is considered to be simultaneously

512 x 512 x 100 x 1 x 1 x 1 x 1 x 1 x 1 x 1 x 1 [...] x 1 %until the 100th
dimension
 | 
Pages: 1
Prev: fzero logical error
Next: arithmetic on ROI in scatter plots