MIJ / Miji and Matlab 2014 a & b

[ericbarnhill]

Miji / MIJ has become a central tool for our lab. There appears to be  
a difference in how Matlab handles MIJ in the 2014 builds. In the 2013  
builds and prior, starting MIJ created a persistent instance that  
could be used for the entire session (notwithstanding unusual errors).  
In 2014, Matlab appears to have thrown a spanner in the works.

Any MIJ instance created only lasts until the next non-MIJ-related  
command. So if I open MIJ, then run my image processing pipeline, the  
instance of MIJ is no longer accessible. I have to open Miji again,  
and it creates a second ImageJ toolbar, etc. Any previous images  
opened in MIJ are no longer accessible and the windows can't even be  
closed only minimised. Eventually memory starts to get sapped by all  
the open dead windows.

It's something of an inconvenience since I would like to keep tweaking  
the pipeline, running it, and comparing the new image to the previous.

I am happy to work on the problem myself if it is still outstanding,  
someone should just get in touch with me about how to contribute.

Eric



---
Eric Barnhill
Clinical Research Imaging Centre
The University of Edinburgh

--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

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

[ctrueden]

Hi Eric,

> I am happy to work on the problem myself if it is still outstanding,
> someone should just get in touch with me about how to contribute.

I hope the lack of response to this thread (now four months old) was not
too discouraging. The thing about MIJ and Miji is: although quite a few
people find it useful, there is no one actively maintaining it. However, as
with the rest of ImageJ and Fiji, there are passive maintainers: people who
will review and merge pull requests, should anyone submit them.

You can find the source for the Miji project at:
https://github.com/fiji/fiji/blob/master/scripts/Miji.m

We do plan to update Miji to use ImageJ2 instead of ImageJ1 [1], but that
work has been put on hold for the time being. One complicating issue is
licensing: Miji is based on MIJ, which has a non-OSS license [2].

One other related project of which you should be aware is the MATLAB script
language for ImageJ's Script Editor. This lets you execute MATLAB code from
within ImageJ (as opposed to calling ImageJ from within MATLAB). For
details, see http://imagej.net/MATLAB

Regards,
Curtis

[1] https://github.com/fiji/fiji/compare/imagej2-matlab
[2] http://bigwww.epfl.ch/sage/soft/mij/#term

On Fri, Dec 19, 2014 at 2:15 AM, Eric Barnhill <[hidden email]>
wrote:

> Miji / MIJ has become a central tool for our lab. There appears to be a
> difference in how Matlab handles MIJ in the 2014 builds. In the 2013 builds
> and prior, starting MIJ created a persistent instance that could be used
> for the entire session (notwithstanding unusual errors). In 2014, Matlab
> appears to have thrown a spanner in the works.
>
> Any MIJ instance created only lasts until the next non-MIJ-related
> command. So if I open MIJ, then run my image processing pipeline, the
> instance of MIJ is no longer accessible. I have to open Miji again, and it
> creates a second ImageJ toolbar, etc. Any previous images opened in MIJ are
> no longer accessible and the windows can't even be closed only minimised.
> Eventually memory starts to get sapped by all the open dead windows.
>
> It's something of an inconvenience since I would like to keep tweaking the
> pipeline, running it, and comparing the new image to the previous.
>
> I am happy to work on the problem myself if it is still outstanding,
> someone should just get in touch with me about how to contribute.
>
> Eric
>
>
>
> ---
> Eric Barnhill
> Clinical Research Imaging Centre
> The University of Edinburgh
>
> --
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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

[ericbarnhill-2]

>Hi Eric, > >> I am happy to work on the problem myself if it is still outstanding, >> someone should just get in touch with me about how to contribute. > >I hope the lack of response to this thread (now four months old) was not >too discouraging. The thing about MIJ and Miji is: although quite a few >people find it useful, there is no one actively maintaining it. However, as >with the rest of ImageJ and Fiji, there are passive maintainers: people who >will review and merge pull requests, should anyone submit them.
Not discouraging but I am glad you posted, as somehow I never managed to find my way to the MIJ repo and I assumed it was a closed-source project for some reason.


The plug-in I hope to release in the next couple weeks converts ImageJ ROIs to Matlab index arrays and vice-versa through the system clipboard. At that time I'll check out the project and see if I can't tweak the code so that Matlab only opens one ImageJ per session. After I work on those two issues I'll submit them for consideration to the repo.


Thanks,
Eric




Sent from [ProtonMail](https://protonmail.ch), encrypted email based in Switzerland.

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

[Dimiter Prodanov-3]

>
>


Hello Eric,

I am glad that you find MIJ useful.
Please also submit the proposed changes to Daniel Sage and myself.
http://bigwww.epfl.ch/sage/soft/mij/

I will be happy also to help.

Dimiter Prodanov

>
> ------------------------------
>
> Date:    Tue, 28 Apr 2015 06:33:38 -0400
> From:    ericbarnhill <[hidden email]>
> Subject: Re: MIJ / Miji and Matlab 2014 a & b
>
> >Hi Eric, > >> I am happy to work on the problem myself if it is still
> outstanding, >> someone should just get in touch with me about how to
> contribute. > >I hope the lack of response to this thread (now four months
> old) was not >too discouraging. The thing about MIJ and Miji is: although
> quite a few >people find it useful, there is no one actively maintaining
> it. However, as >with the rest of ImageJ and Fiji, there are passive
> maintainers: people who >will review and merge pull requests, should anyone
> submit them.
> Not discouraging but I am glad you posted, as somehow I never managed to
> find my way to the MIJ repo and I assumed it was a closed-source project
> for some reason.
>
>
> The plug-in I hope to release in the next couple weeks converts ImageJ
> ROIs to Matlab index arrays and vice-versa through the system clipboard. At
> that time I'll check out the project and see if I can't tweak the code so
> that Matlab only opens one ImageJ per session. After I work on those two
> issues I'll submit them for consideration to the repo.
>
>
> Thanks,
> Eric
>
>
>
> *************************************************************
>

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