moving boundary

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|

moving boundary

WilliamBWright
I have a few thousand images documenting the movement of a boundary (ice growth on a surface). When I play the images in an animation, I can see the boundary moving. I want to extract the location of the boundary and get a table of thickness versus time. It seems to me that the program should be able to do this easily but I have not been able to pick the right function after randomly trying several of the built-in options. I haven't used this program before and I would really appreciate some tips. Thanks

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: moving boundary

Rob van 't Hof-2
Hi William,
Giving us a few sample images will be very helpful. Coming up with a
strategy without knowing what the images look like is pretty much
impossible.
bye,
rob

On 16/10/2012 22:59, William Wright wrote:
> I have a few thousand images documenting the movement of a boundary (ice growth on a surface). When I play the images in an animation, I can see the boundary moving. I want to extract the location of the boundary and get a table of thickness versus time. It seems to me that the program should be able to do this easily but I have not been able to pick the right function after randomly trying several of the built-in options. I haven't used this program before and I would really appreciate some tips. Thanks
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
_____________________________
Dr. Rob van 't Hof
Reader

Centre for Molecular Medicine
MRC IGMM
University of Edinburgh
Western General Hospital
Crewe Road, Edinburgh EH4 2XU
United Kingdom


Phone: (+44)-131-6511031
email: [hidden email]
_____________________________


The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: moving boundary

WilliamBWright
In reply to this post by WilliamBWright
Here's a sample image There's about 15000 images for this case. The boundary can be seen on the right. There is good contrast between the boundary and the background although on the left there is a reflection off the geometry

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

Run 572 Cam 3 00001.jpg (521K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: moving boundary

Gabriel Landini
On Thursday 18 Oct 2012 16:22:03 William Wright  wrote:
> Here's a sample image There's about 15000 images for this case. The boundary
> can be seen on the right. There is good contrast between the boundary and
> the background although on the left there is a reflection off the geometry

Hi,
It might be contrasted in that shot, but the background is very unevenly
illuminated and so it might be difficult to make the boundary as it grows to
the left part of the image which is much brighter.

If this is a sequence, have you thought of computing the differences between
consecutive shots?

Another issue would be to try to remove the background if it remains constant
(for example subtracting or ratio-ing every shot with the first shot in the
series).
Hope it helps.

Regards

Gabriel

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: moving boundary

Kenton Arkill
In reply to this post by WilliamBWright
Hi
Assuming that the big bright line is the boundary then you could either use a threshold or simply an xy profile to find the point (and find the peak in the smoothed second differential… might not have to smooth). The cycle through all the images and plot the x coordinate…
Regards
Kenton

Kenton Arkill
GB U23 Ladies'
Underwater Hockey Coach
GB Masters
Underwater Hockey Manager




On 18 Oct 2012, at 17:22, William Wright wrote:

> Here's a sample image There's about 15000 images for this case. The boundary can be seen on the right. There is good contrast between the boundary and the background although on the left there is a reflection off the geometry
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> <Run 572 Cam 3 00001.jpg>


--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: moving boundary

Gabriel Landini
In reply to this post by WilliamBWright
On Thursday 18 Oct 2012 16:22:03 you wrote:
> Here's a sample image There's about 15000 images for this case. The boundary
> can be seen on the right. There is good contrast between the boundary and
> the background although on the left there is a reflection off the geometry

Hm... is the boundary the vertical line, or the white blob in the right?
My previous comment about background subtraction was about the blob, but now I
realise that it might be the vertical edge (which would be much easier to
segment).
Regards

Gabriel

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: moving boundary

WilliamBWright
In reply to this post by WilliamBWright
Here is an update of where I am at: If I set the threshold to about 80 [setThreshold(0,80)] it seems to give me the image I want. I ran the commands

run("Input/Output...", :jpeg=80 gif=-1 file=.txt copy_row save_column save_row");
run("Image to Results");

I get a table listing values of 0 or 255 showing me where the boundary is located. What I want to do is to look at the middle 10 pixels (there's 720 in the y-direction, so let's say pixel 355 to 365 in the y-direction) and then find the x-location where it changes from 0 to 255 and then output that value. For the image I tested, that would be pixel 1073 in the x-direction. For 1 image I can manually look at the table, but for 15000 images per run I would like something more automated. Thank you very much so far for your help

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: moving boundary

Jerome Mutterer-3
Hi,
you could try pasting the following macro in the batch processor and apply
it to a folder containing your 15000 images.

t=getTitle;
run("Convolve...", "text1=[1 0 -1\n] normalize");
makeRectangle(0, (getHeight-10)/2, getWidth, 10);
run("Reslice [/]...", "output=1.000 start=Top avoid");
run("Z Project...", "start=1 stop=10 projection=[Average Intensity]");
run("Find Maxima...", "noise=100 output=[Point Selection]");
if (selectionType==10) {
getSelectionCoordinates(x,y);
if (x.length==1) print (t+":"+x[0]);
}
close(); // the Z-projection
close(); // the reslice
close(); // the original image
// end of macro


Sincerely,

Jerome



On 24 October 2012 21:58, William Wright <[hidden email]> wrote:

> Here is an update of where I am at: If I set the threshold to about 80
> [setThreshold(0,80)] it seems to give me the image I want. I ran the
> commands
>
> run("Input/Output...", :jpeg=80 gif=-1 file=.txt copy_row save_column
> save_row");
> run("Image to Results");
>
> I get a table listing values of 0 or 255 showing me where the boundary is
> located. What I want to do is to look at the middle 10 pixels (there's 720
> in the y-direction, so let's say pixel 355 to 365 in the y-direction) and
> then find the x-location where it changes from 0 to 255 and then output
> that value. For the image I tested, that would be pixel 1073 in the
> x-direction. For 1 image I can manually look at the table, but for 15000
> images per run I would like something more automated. Thank you very much
> so far for your help
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: moving boundary

WilliamBWright
In reply to this post by WilliamBWright
Is there documentation for these commands? I would really like to understand what is going on when I run this macro

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html