Increase Class in Trainable WEKA Segmentation

[Charalampos Zachariadis]

Hello All,

My question as the topic suggests is that I am wondering how can I increase the trainable WEKA segmentation plugin classes over the prefixed maximum of 10. The reason is that I am trying to classify some RapidEye imagery for vegetation changes detection and the vegetation classes in my area exceeds the maximum allowed.

Kind Regards,
Charalampos

PGT Student at the University of Glasgow

[Ignacio Arganda-Carreras-2]

Dear Charalampos,

Do you need to use the GUI of the plugin for this? In that case, you will
have to recompile the code after increasing the constant value from 10 to
the number you need:

https://github.com/fiji/Trainable_Segmentation/blob/master/src/main/java/trainableSegmentation/WekaSegmentation.java#L98

If you're working from a script, you may have other solutions.

ignacio

On Fri, Jul 29, 2016 at 11:20 AM, Charalampos Zachariadis <
[hidden email]> wrote:

> Hello All,
>
> My question as the topic suggests is that I am wondering how can I increase
> the trainable WEKA segmentation plugin classes over the prefixed maximum of
> 10. The reason is that I am trying to classify some RapidEye imagery for
> vegetation changes detection and the vegetation classes in my area exceeds
> the maximum allowed.
>
> Kind Regards,
> Charalampos
>
> PGT Student at the University of Glasgow
>
>
>
>
> --
> View this message in context:
> http://imagej.1557.x6.nabble.com/Increase-Class-in-Trainable-WEKA-Segmentation-tp5016951.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
Ignacio Arganda-Carreras, Ph.D.
Ikerbasque Research Fellow
Departamento de Ciencia de la Computacion e Inteligencia Artificial
Facultad de Informatica, Universidad del Pais Vasco
Paseo de Manuel Lardizabal, 1
20018 Donostia-San Sebastian
Guipuzcoa, Spain

Phone : [phone]
Website: http://sites.google.com/site/iargandacarreras/
<http://biocomp.cnb.csic.es/~iarganda/index_EN.html>

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

[Charalampos Zachariadis]

Dear Dr. Ignacio,

Thank you for your quick response, but could you please explain to me the whole procedure cause I do not have the experience since it is my first time using a machine learning platform to do it and I have been trying since I got your answer. I apologize for this and I thank you in advance.

Kind regards,
Charalampos

[Ignacio Arganda-Carreras-2]

Don't worry, I will change that limit in the plugin and release a new
version with it soon.

In the meantime, can you show an image that you're trying to segment? Do
you have more than 10 types of vegetation?

On Fri, Jul 29, 2016 at 11:14 PM, Charalampos Zachariadis <
[hidden email]> wrote:

> Dear Dr. Ignacio,
>
> Thank you for your quick response, but could you please explain to me the
> whole procedure cause I do not have the experience since it is my first
> time
> using a machine learning platform to do it and I have been trying since I
> got your answer. I apologize for this and I thank you in advance.
>
> Kind regards,
> Charalampos
>
>
>
> --
> View this message in context:
> http://imagej.1557.x6.nabble.com/Increase-Class-in-Trainable-WEKA-Segmentation-tp5016951p5016954.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
Ignacio Arganda-Carreras, Ph.D.
Ikerbasque Research Fellow
Departamento de Ciencia de la Computacion e Inteligencia Artificial
Facultad de Informatica, Universidad del Pais Vasco
Paseo de Manuel Lardizabal, 1
20018 Donostia-San Sebastian
Guipuzcoa, Spain

Phone : [phone]
Website: http://sites.google.com/site/iargandacarreras/
<http://biocomp.cnb.csic.es/~iarganda/index_EN.html>

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

[Charalampos Zachariadis]

Dear Dr. Ignacio,

As you can see it is quite a large area (Ox Mountains, Ireland) and as per the classes the two pictures present some of them, as the ocean will be added, cities, roads, agriculture areas. I need to prepare one image and train the model and then apply it to another 3 images from previous years and one from 2015 in order to present the changes in vegetation over the time.

Thanks again for everything,
Charalampos

[Charalampos Zachariadis]

Dear Dr. Ignacio,

I managed to do and increase the classes and now I am happily working with the Weka Segmentation. Thank you for all your time and hard work.

Kind Regards,
Charalampos

[Ignacio Arganda-Carreras-2]

Dear Charlampos,

I have just made a new release with a maximum number of classes of 100.
Please go ahead and update your Fiji to get the latest version of the
plugin.

Best regards,

ignacio

On Wed, Aug 3, 2016 at 10:30 PM, Charalampos Zachariadis <
[hidden email]> wrote:

> Dear Dr. Ignacio,
>
> I managed to do and increase the classes and now I am happily working with
> the Weka Segmentation. Thank you for all your time and hard work.
>
> Kind Regards,
> Charalampos
>
>
>
> --
> View this message in context:
> http://imagej.1557.x6.nabble.com/Increase-Class-in-Trainable-WEKA-Segmentation-tp5016951p5016971.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
Ignacio Arganda-Carreras, Ph.D.
Ikerbasque Research Fellow
Departamento de Ciencia de la Computacion e Inteligencia Artificial
Facultad de Informatica, Universidad del Pais Vasco
Paseo de Manuel Lardizabal, 1
20018 Donostia-San Sebastian
Guipuzcoa, Spain

Phone : [phone]
Website: http://sites.google.com/site/iargandacarreras/
<http://biocomp.cnb.csic.es/~iarganda/index_EN.html>

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