[GoFigure2-developers] GoFigure2-v0.8.1 tagged!

[Arnaud GELAS]

Hi all,

All major and critical bugs from GoFigure2-v0.8.0 have been fixed, so we
have just created a new tag.

Important notice, this release will be the last release before the
introduction of the SuperBuild feature which will require CMake 2.8.3 to
fully benefit from it.
This release is also one of the last one which won't require an ITK4
version (even though ITK4 is in alpha).

This upcoming change won't affect the development of GoFigure2, but it
will for packages...

Best,
Arnaud

------------------------------------------------------------------------------
Enable your software for Intel(R) Active Management Technology to meet the
growing manageability and security demands of your customers. Businesses
are taking advantage of Intel(R) vPro (TM) technology - will your software
be a part of the solution? Download the Intel(R) Manageability Checker
today! http://p.sf.net/sfu/intel-dev2devmar
_______________________________________________
GoFigure2-developers mailing list
GoFigure2-...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/gofigure2-developers

[Antonin Perrot-Audet]

Hi !

I developped a pipeline for large images processing, and especially,
nuclei segmentation. There is still some parameter tuning to do, but it
is working nicely.
It is now coded as a main program, and each box is an itk filter. I
joined a schematic of the pipeline. Can you guys (Nico? Arnaud?) tell me
under which form you need it ?

I was thinking of something that could be amazing : as all my filters
are 2D/3D, do you think it would be possible for the user to select a
2D(3D?) region, and send it to the pipeline, to test it, and see the
results quickly, before starting the process on a big (big big) image ?

About mutlithreading : that would be cool to start one process per
subimage, and force the part if the pipeline dealing with subimages to
be monothreaded, so we would process up to N nucleis at the time... but
it is not done yet.

That would also be cool to be able to change part of the pipeline
dynamically (nuclei detection/nuclei segmentation).

--
Antonin Perrot-Audet
Electrical Engineering& Computer Sciences, INSA Lyon
M2 GEGP Signal& Image Processing, INSA Lyon, UCBL, Centrale Lyon

[Arnaud GELAS]

Hi Antonin,

On Mar 31, 2011, at 5:50 AM, Antonin Perrot-Audet wrote:

Hi !

I developped a pipeline for large images processing, and especially,
nuclei segmentation. There is still some parameter tuning to do, but it
is working nicely.

Excellent!

It is now coded as a main program, and each box is an itk filter. I
joined a schematic of the pipeline. Can you guys (Nico? Arnaud?) tell me
under which form you need it ?

The first things (on your side, if you want your code to be integrated as it 

is in GoFigure2):

- Are you allowed to release this source code?

- What is the license of this source code?

Depending on the answers to these questions, there are different ways.

Let's assume everything works fine!

Since your pipeline is entirely performed in ITK, what about submitting

an article to the Insight Journal? (The code can stay on github for instance,

and be reviewed by experts).

Lydie is currently working on an easy way to integrate new algorithm, where

you only need to implement couples of methods (for the time being, then later

you would only need to provide the algorithm code and one xml file).

Lydie can guide you!

I was thinking of something that could be amazing : as all my filters
are 2D/3D, do you think it would be possible for the user to select a
2D(3D?) region, and send it to the pipeline, to test it, and see the
results quickly, before starting the process on a big (big big) image ?

The first thing we would integrate into GoFigure2 is one segmentation
procedure where the user would click
    - either the center of a cell and a radius, then the cell would be
segmented and extracted (this would by-pass your seed selection process)
    - either a point in the image with a radius, then your would pipeline would
work on a ROI.

Then we could work on something funkier, and nicer as you mention!

About mutlithreading : that would be cool to start one process per
subimage, and force the part if the pipeline dealing with subimages to
be monothreaded, so we would process up to N nucleis at the time... but
it is not done yet.

That would also be cool to be able to change part of the pipeline
dynamically (nuclei detection/nuclei segmentation).

Best,
Arnaud

--
Antonin Perrot-Audet
Electrical Engineering&  Computer Sciences, INSA Lyon
M2 GEGP Signal&  Image Processing, INSA Lyon, UCBL, Centrale Lyon

<CellDetectionPipe.pdf><ATT00001.txt><ATT00002.txt>