wish list item

[Aryeh Weiss]

I see that the particle analyzer can filter on size and circulatrity.
It would be nice if it were possible to filter on any of the measured
parameters.
One way to do it might be to extend the set measurements menu to allow
upper and lower limits to be entered next to each measurable parameter.

Does such a tool exist? I kow one can read the results table and  loop
though to keep or discard objects, but I was wondering if there is
already something that does this, which maybe I missed.

--aryeh
--
Aryeh Weiss
School of Engineering
Bar Ilan University
Ramat Gan 52900 Israel

Ph:  972-3-5317638
FAX: [phone]

[Gabriel Landini]

On Tuesday 06 July 2010 21:08:03 you wrote:
> I see that the particle analyzer can filter on size and circulatrity.
> It would be nice if it were possible to filter on any of the measured
> parameters.

Maybe this would be useful
http://www.dentistry.bham.ac.uk/landinig/software/classify/classify.html

Cheers

Gabriel

[GCH-2]

Try the Shape descriptor.

http://www.gcsca.net/IJ/Shapes.html

Gary.

On Jul 6, 2010, at 10:28 PM, Gabriel Landini wrote:

> On Tuesday 06 July 2010 21:08:03 you wrote:
>> I see that the particle analyzer can filter on size and circulatrity.
>> It would be nice if it were possible to filter on any of the measured
>> parameters.
>
> Maybe this would be useful
> http://www.dentistry.bham.ac.uk/landinig/software/classify/ 
> classify.html
>
> Cheers
>
> Gabriel

[Aryeh Weiss]

GCH wrote:

> Try the Shape descriptor.
>
> http://www.gcsca.net/IJ/Shapes.html
>
> Gary.
>
> On Jul 6, 2010, at 10:28 PM, Gabriel Landini wrote:
>
>> On Tuesday 06 July 2010 21:08:03 you wrote:
>>> I see that the particle analyzer can filter on size and circulatrity.
>>> It would be nice if it were possible to filter on any of the measured
>>> parameters.
>>
>> Maybe this would be useful
>> http://www.dentistry.bham.ac.uk/landinig/software/classify/classify.html
>>
>> Cheers
>>
>> Gabriel
>

Thank you Gary and Gabriel. Both of these plugins can do what I need. I
have had a few problems and comments, with which perhaps you can assist,
and may happen to others on the list.

Gary's plugin is limited to shape descriptors (which is fine -- that is
the name and purpose of the plugin). I wonder if much of its
functionality can be integrated into Analyze>Set Measurements ? This
would provide a general way filter on any item that is measured.
One other thing -- I could only get Gary's plugin to work with dark
objects on a light background. I had dark background selected in the
threshold interface, and also I have the background color set to black
and foreground to white in the options menu. I must have missed a
different place where this much be set.

Gabriel's plugin can select anything in the results table, which is
nice, but if you need a range (ie low_limit<parameter<high_limit) then
you  are limited to two parameters (please correct me if I
misunderstood). However, I am having trouble running it.

I downloaded the Morphology set and it resides in
$FIJI_HOME/plugins/Morphology,
When I try to run it, I get:

java.lang.NoClassDefFoundError: Morphology/Classify_Particles (wrong
name: Classify_Particles)
        at java.lang.ClassLoader.defineClass1(Native Method)

This usually means that the plugin expects to be in a different
directory (but other plugins in the set run fine).

I can compile and run, in which case the plugin runs, and shows me the
interface. However, following "ok". I get this message:

Plugin or class not found: "scripting.java.Refresh_Javas"
(java.lang.ClassNoFOundException: scripting.java.Refresh_Javas)

So here I got stuck. Before I try dumping all the class files into the
top level plugins directory, I was hoping that the many people more
knowledgeable in Java than I would be able to provide a solution.

Some other details: I run Fiji, but the ij.jar is 1.44d16. OS is OSX 10.5.8.

I have not yet had a chance to check this, but are these plugins
"macro-friendly"?

Thanks again to both Gary and Gabriel for a very nice set of tools that
they provided.

--aryeh
--
Aryeh Weiss
School of Engineering
Bar Ilan University
Ramat Gan 52900 Israel

Ph:  972-3-5317638
FAX: [phone]

[Gabriel Landini]

On Wednesday 07 Jul 2010  11:26:56 Aryeh Weiss wrote:
> Gabriel's plugin can select anything in the results table, which is
> nice, but if you need a range (ie low_limit<parameter<high_limit) then
> you  are limited to two parameters (please correct me if I
> misunderstood).

Yes, 2 if you want 2 thresholds, 4 if you want one threshold per parameter.
You could add more if you wanted... you have the source. Originally I did not
put more because the dialog was getting ridiculously wrong. Maybe this can be
arranged better.

Also there is nothing stopping running the same plugin more than once, so you
can make more complex selections than "all AND" or "all OR".

> However, I am having trouble running it.
>
> I downloaded the Morphology set and it resides in
> $FIJI_HOME/plugins/Morphology,
> When I try to run it, I get:
>
> java.lang.NoClassDefFoundError: Morphology/Classify_Particles (wrong
> name: Classify_Particles)
> at java.lang.ClassLoader.defineClass1(Native Method)
>
> This usually means that the plugin expects to be in a different
> directory (but other plugins in the set run fine).

This problem is, I think, due to Fiji expecting every plugin to be in a jar
with a configuration file (unlike ImageJ, where you can have the java files
together with the class files).

I know that Johannes is aware of this, so it might get fixed sometime soon.

In the mean time, just move the *.java files out of the plugins/Morphology
folder and restart Fiji (it works fine here after doing this).

> I have not yet had a chance to check this, but are these plugins
> "macro-friendly"?

They should be if they show a dialog. Let me know if they do not.

Cheers
G.

[Gabriel Landini]

On Wednesday 07 Jul 2010  11:41:01 I wrote:
>
> Originally I did  not put more because the dialog was getting ridiculously
> wrong.

I meant "long" not "wrong" :-)

Cheers

G.