Rank: 829 based on 158 downloads (last 30 days) and 5 files submitted
photo

Joshua Stough

E-mail
Company/University
Washington and Lee University

Personal Profile:
Professional Interests:
image processing, medical image analysis, computer vision

 

Watch this Author's files

 

Files Posted by Joshua View all
Updated   File Tags Downloads
(last 30 days)
Comments Rating
23 Apr 2013 Screenshot 4D Volume Visualization 4D (3D+1) volume visualization with windowing, custom colormaps, and component selection Author: Joshua Stough 4d, vectorvalued volumetr..., visualization, data exploration, gui 38 4
  • 4.0
4.0 | 1 rating
23 Apr 2013 Screenshot 3D Volume Visualization 3D volume viz with interactive slice selection, colored ortho-planes, windowing, colormap selection. Author: Joshua Stough 3d visualization, gui, slice, orthogonal planar vie... 53 2
29 Aug 2012 Screenshot Select Data Select original data through a projective plot of that data. Author: Joshua Stough data exploration, gui, demo 17 0
01 Aug 2012 Screenshot Random Sample from Discrete PDF Inverse transform sampling to generate random sample from pdf given by domain x and p(x). Author: Joshua Stough statistics, demo, data exploration, inverse transform sam... 35 3
  • 5.0
5.0 | 1 rating
19 Jun 2012 Screenshot Photo Quadrat Selection Tool An interactive tool for manually selecting photo quadrats from a collection of images. Author: Joshua Stough image processing, gui, quadrat selection, homography 15 0
Comments and Ratings by Joshua View all
Updated File Comments Rating
01 May 2014 4D Volume Visualization 4D (3D+1) volume visualization with windowing, custom colormaps, and component selection Author: Joshua Stough

Response to John:
Thank you for your question John. Your question concerned why certain slicing of your data seem to appear in the wrong orthoplanes. While I am slightly chagrined, the honest response is that I have intentionally permuted and flipdim'd to make my particular data (MRI images in analyze75 format) appear as I want it (RAS). Search for 'permute' and 'flipdim' in the code to customize it for yourself.
As an aside, the particular examples you provide disagree with your own description. For example, Aq(c,:,:,f) is not plotting the xz plane, it's actually the yz plane (all y and all z for a particular x).

15 Dec 2013 Random Sample from Discrete PDF Inverse transform sampling to generate random sample from pdf given by domain x and p(x). Author: Joshua Stough

Response to Chavas:
Thank you so much for your comment. The problem as I see it is that with the assumption of 0, it is possible to generate samples from outside of the provided domain x. Even for the examples I provide, cdf(1) > 0 and so invalid samples can be generated. However, your proposed solution could also generate such samples. Perhaps to account for binning in x, x(1) - dx/2 should be possible as a minimum, and x(end) + dx/2 as a max (assuming equal sampling of the domain). The other option is to assume the user provides precisely the x(1) and x(end) they want to sample within. I will post new code soon. Thank you again.

06 Jul 2013 3D Volume Visualization 3D volume viz with interactive slice selection, colored ortho-planes, windowing, colormap selection. Author: Joshua Stough

Response to Gustavo: My code doesn't care how you got the image. If it, say 'I', exists in matlab as a 3d array of any class (use size and class functions to determine this), then just try 'vis3d(double(I));' instead of whatever you were doing.

For loading nifti images, exchange #8797 seems popular: http://www.mathworks.com/matlabcentral/fileexchange/8797-tools-for-nifti-and-analyze-image

Good luck,
stough

06 Jul 2013 4D Volume Visualization 4D (3D+1) volume visualization with windowing, custom colormaps, and component selection Author: Joshua Stough

Response to Gustavo: My code doesn't care how you got the image. If it, say 'I', exists in matlab as a 4d array of any class (use size and class functions to determine this), then just try 'vis4d(double(I));' instead of whatever you were doing.

For loading nifti images, exchange #8797 seems popular: http://www.mathworks.com/matlabcentral/fileexchange/8797-tools-for-nifti-and-analyze-image

Good luck,
stough

19 Feb 2013 M-code LaTeX Package Easily include nicely syntax highlighted m-code in your LaTeX documents. Author: Florian Knorn

Worked great out of the box. Thank you

Comments and Ratings on Joshua's Files View all
Updated File Comment by Comments Rating
01 May 2014 4D Volume Visualization 4D (3D+1) volume visualization with windowing, custom colormaps, and component selection Author: Joshua Stough Stough, Joshua

Response to John:
Thank you for your question John. Your question concerned why certain slicing of your data seem to appear in the wrong orthoplanes. While I am slightly chagrined, the honest response is that I have intentionally permuted and flipdim'd to make my particular data (MRI images in analyze75 format) appear as I want it (RAS). Search for 'permute' and 'flipdim' in the code to customize it for yourself.
As an aside, the particular examples you provide disagree with your own description. For example, Aq(c,:,:,f) is not plotting the xz plane, it's actually the yz plane (all y and all z for a particular x).

31 Mar 2014 4D Volume Visualization 4D (3D+1) volume visualization with windowing, custom colormaps, and component selection Author: Joshua Stough John

Hi Joshua,
thanks for your software. I have found it very useful.
I wondered if you would be able to explain something to me. If I plot Aq(constant,:,:,frame), I am plotting the xz plane. When I visualize it with your software, I see this cut in the yz plane. (similar occurrence with Aq(:,constant,:,frame, this time the yz cut appears in xz.) Is there any reason for this, or have I misunderstood something?
thanks,
John

25 Mar 2014 4D Volume Visualization 4D (3D+1) volume visualization with windowing, custom colormaps, and component selection Author: Joshua Stough Mihai

15 Dec 2013 Random Sample from Discrete PDF Inverse transform sampling to generate random sample from pdf given by domain x and p(x). Author: Joshua Stough Stough, Joshua

Response to Chavas:
Thank you so much for your comment. The problem as I see it is that with the assumption of 0, it is possible to generate samples from outside of the provided domain x. Even for the examples I provide, cdf(1) > 0 and so invalid samples can be generated. However, your proposed solution could also generate such samples. Perhaps to account for binning in x, x(1) - dx/2 should be possible as a minimum, and x(end) + dx/2 as a max (assuming equal sampling of the domain). The other option is to assume the user provides precisely the x(1) and x(end) they want to sample within. I will post new code soon. Thank you again.

13 Dec 2013 Random Sample from Discrete PDF Inverse transform sampling to generate random sample from pdf given by domain x and p(x). Author: Joshua Stough Chavas, Dan

Whoops, there is a problem: if you input a pdf (px) such that px(1)>0, then cdf(1)>0. However it is possible for rnd < cdf(1) ; in such a case, interp1 returns NaN, which you've forced to be 0. This is wrong.

Consider adding this code to ensure the cdf begins at 0:

%%Account for non-zero probability at start
cdf = cumsum(px);
if(cdf(1)>0)
cdf = [0 cdf];
dx = x(2) - x(1);
x0 = x(1) - dx;
x = [x0 x];
end

rnd = rand(sampleSize, 1);

X = interp1(cdf, x, rnd, 'linear');

Contact us