Help concerning Temperature Map

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

Help concerning Temperature Map

jgodeke
Hello!

I´ve spend some time with imagej in the last weeks because i have a difficult task to solve.

Now that i have some results, i´m not that happy with i would like to know if someone knows a better solution.
original image
This is the source temperature map. Aim is to segment connected areas with the same/similar temperature to define the average size. Actual temperature of a "cell" is not of concern.
This is my solution:
my solution
edit:(I despeckled the image)
I selected red green and blue areas via color thresholding separately and used watersheding to receive the outlines which i copied over the actual image in photoshop. The average area is 182 pixels which "should" be rather something like 120 pixels. The lake in the top middle-left is segmented though i don´t know why and some other areas too.

I would be very happy if someone knows a plugin like watershed which produces better results!

Thank you,

Jonathan
Reply | Threaded
Open this post in threaded view
|

Re: Help concerning Temperature Map

David Webster
I'm curious, does an actual original temperture map exist or onlyu the
psuedo colored image.

David Webster

On Fri, May 6, 2011 at 9:45 AM, jgodeke <[hidden email]> wrote:

> Hello!
>
> I´ve spend some time with imagej in the last weeks because i have a
> difficult task to solve.
>
> Now that i have some results, i´m not that happy with i would like to know
> if someone knows a better solution.
> http://imagej.588099.n2.nabble.com/file/n6338176/temp.jpg
> This is the source temperature map. Aim is to segment connected areas with
> the same/similar temperature to define the average size. Actual temperature
> of a "cell" is not of concern.
> This is my solution:
> http://imagej.588099.n2.nabble.com/file/n6338176/rotegrueneblaue.jpg
> I selected red green and blue areas via color thresholding separately and
> used watersheding to receive the outlines which i copied over the actual
> image in photoshop. The average area is 182 pixels which "should" be rather
> something like 120 pixels. The lake in the top middle-left is segmented
> though i don´t know why and some other areas too.
>
> I would be very happy if someone knows a plugin like watershed which
> produces better results!
>
> Thank you,
>
> Jonathan
>
>
> --
> View this message in context:
> http://imagej.588099.n2.nabble.com/Help-concerning-Temperature-Map-tp6338176p6338176.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
Reply | Threaded
Open this post in threaded view
|

Re: Help concerning Temperature Map

jgodeke
If an actual temperature map would be greyscale then it exists. Somehow i never considered working with it.
Reply | Threaded
Open this post in threaded view
|

Re: Help concerning Temperature Map

David Webster
Well, now that I thin about it, segmenting the psuedo color map may easier
given ImageJ's available tools.

I amn curious how you computed the aevrages  Did you extract the areas for
each cell, and then compute their mean value. I so were you able to look for
outlier  regions with areas >> 120, Can you your cpompue the region median
in addition to mean?

David Webster

On Fri, May 6, 2011 at 11:11 AM, jgodeke <[hidden email]> wrote:

> If an actual temperature map would be greyscale then it exists. Somehow i
> never considered working with it.
>
> --
> View this message in context:
> http://imagej.588099.n2.nabble.com/Help-concerning-Temperature-Map-tp6338176p6338428.html
>  Sent from the ImageJ mailing list archive at Nabble.com.
>
Reply | Threaded
Open this post in threaded view
|

Re: Help concerning Temperature Map

chesit