Login  Register

Thresholding problem

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
3 messages Options Options
Embed post
Permalink
Reply | Threaded
Open this post in threaded view
| More
Print post
Permalink

Thresholding problem

Christian Liebig-2
7 posts
Dear all,

I use Objectcounter 3D to find objects in my image. As output I get an
image with all objects coded in  a different intensity value (sample
picture attached). Now I want to threshold the image that I get only a
single object. If I do that using the threshold function of ImageJ
(adjusting upper and lower threshold to e.g. 265) I very often get 2
objects thresholded. Is that a bug or how can I threshold for only a
single intensity value?

Thanks in advance and kind regards,

Christian


--
Christian Liebig, PhD

Light Microscopy Facility
Max-Planck-Institute for Developmental Biology
Spemannstrasse 35
D-72076 Tuebingen
Germany

Phone:  +49 7071 601 443
e-mail: [hidden email]
e-mail2: [hidden email]

--
Christian Liebig, PhD

Light Microscopy Facility
Max-Planck-Institute for Developmental Biology
Spemannstrasse 35
D-72076 Tuebingen
Germany

Phone:  +49 7071 601 443
e-mail: [hidden email]
e-mail2: [hidden email]



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

Particles.tif (963K) Download Attachment
Reply | Threaded
Open this post in threaded view
| More
Print post
Permalink

Re: Thresholding problem

Gabriel Landini
1783 posts
On Wednesday 21 Nov 2012 16:22:45 you wrote:
> I use Objectcounter 3D to find objects in my image. As output I get an
> image with all objects coded in  a different intensity value (sample
> picture attached). Now I want to threshold the image that I get only a
> single object. If I do that using the threshold function of ImageJ
> (adjusting upper and lower threshold to e.g. 265) I very often get 2
> objects thresholded. Is that a bug or how can I threshold for only a
> single intensity value?

If you rimage has more than 256 levels, then the image histogram is mapped to
256, so I guess that two close values are binned to the same greyscale level.
If you put the cursor on top of those objects thresholded at 265 you will see
that one has grey level 264 and the other 265.

To threshold them, set the threshold in the Thresholder applet with the "Set"
button (eg min 264 and max 264) and then run the command "Make Binary". Only
one of the two objects (the one painted in 264) is thresholded.

Cheers
Gabriel

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

Re: Thresholding problem

Christian Liebig-2
7 posts
Hi,
also works with "Convert to Mask" and "Create Selection", which is sort
of giving me what I need.
Great!
Many thanks,
Christian

On 21.11.2012 17:53, Gabriel Landini wrote:

> On Wednesday 21 Nov 2012 16:22:45 you wrote:
>> I use Objectcounter 3D to find objects in my image. As output I get an
>> image with all objects coded in  a different intensity value (sample
>> picture attached). Now I want to threshold the image that I get only a
>> single object. If I do that using the threshold function of ImageJ
>> (adjusting upper and lower threshold to e.g. 265) I very often get 2
>> objects thresholded. Is that a bug or how can I threshold for only a
>> single intensity value?
> If you rimage has more than 256 levels, then the image histogram is mapped to
> 256, so I guess that two close values are binned to the same greyscale level.
> If you put the cursor on top of those objects thresholded at 265 you will see
> that one has grey level 264 and the other 265.
>
> To threshold them, set the threshold in the Thresholder applet with the "Set"
> button (eg min 264 and max 264) and then run the command "Make Binary". Only
> one of the two objects (the one painted in 264) is thresholded.
>
> Cheers
> Gabriel
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html


--
Christian Liebig, PhD

Light Microscopy Facility
Max-Planck-Institute for Developmental Biology
Spemannstrasse 35
D-72076 Tuebingen
Germany

Phone:  +49 7071 601 443
e-mail: [hidden email]
e-mail2: [hidden email]

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