last release of HistThresh

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

last release of HistThresh

Carnë Draug
Hi Barry

early last year you have contacted Antti Niemistö about his HistThresh
toolbox [1]. This was a Matlab toolbox for image thresholding which he
released under GPLv2+. In the mailing list, he agreed to release it
under BSD for ImageJ2 development.

Unfortunately, it appears that he no longer works at the Tampere
University of Technology and the website for the project has been
taken down. His previous email address has also been cancelled and I
couldn't find his new contact. I tried to use wayback machine to
salvage his project, but the last backup [2] was before the last
release (he made one release prompted by your email which fixed a
couple of bugs in the minimum and maximum likelihood method). I have
made a github repository with the version I found [3]

I made an Octave port which works fine for me [4] but I know some
Matlab users who would like to use it. Would you happen to have made a
backup of the last release he made?

Thanks in advance,
Carnë

[1] http://imagej.net/pipermail/imagej-devel/2013-February/001404.html
[2] http://web.archive.org/web/20120628113512/http://www.cs.tut.fi/~ant/histthresh/
[3] https://github.com/carandraug/histthresh
[4] https://sourceforge.net/p/octave/image/ci/default/tree/inst/graythresh.m

_______________________________________________
ImageJ-devel mailing list
[hidden email]
http://imagej.net/mailman/listinfo/imagej-devel
Reply | Threaded
Open this post in threaded view
|

Re: last release of HistThresh

Barry DeZonia
Hi Carnë,

I am no longer working with the lab that worked with Antti. I have looked at the backups I made when I left and I do not have a copy of Antti's work. Some people in my old lab might have it and they check this mailing list. However I have also CCed Gabriel Landini (affiliated with Fiji) as he worked with me on this a bit and may have made a copy of the code.


On Tue, Mar 18, 2014 at 5:04 PM, Carnë Draug <[hidden email]> wrote:
Hi Barry

early last year you have contacted Antti Niemistö about his HistThresh
toolbox [1]. This was a Matlab toolbox for image thresholding which he
released under GPLv2+. In the mailing list, he agreed to release it
under BSD for ImageJ2 development.

Unfortunately, it appears that he no longer works at the Tampere
University of Technology and the website for the project has been
taken down. His previous email address has also been cancelled and I
couldn't find his new contact. I tried to use wayback machine to
salvage his project, but the last backup [2] was before the last
release (he made one release prompted by your email which fixed a
couple of bugs in the minimum and maximum likelihood method). I have
made a github repository with the version I found [3]

I made an Octave port which works fine for me [4] but I know some
Matlab users who would like to use it. Would you happen to have made a
backup of the last release he made?

Thanks in advance,
Carnë

[1] http://imagej.net/pipermail/imagej-devel/2013-February/001404.html
[2] http://web.archive.org/web/20120628113512/http://www.cs.tut.fi/~ant/histthresh/
[3] https://github.com/carandraug/histthresh
[4] https://sourceforge.net/p/octave/image/ci/default/tree/inst/graythresh.m


_______________________________________________
ImageJ-devel mailing list
[hidden email]
http://imagej.net/mailman/listinfo/imagej-devel
Reply | Threaded
Open this post in threaded view
|

Re: last release of HistThresh

Gabriel Landini
On Tuesday 18 Mar 2014 18:31:58 Barry DeZonia wrote:

> I am no longer working with the lab that worked with Antti. I have looked
> at the backups I made when I left and I do not have a copy of Antti's work.
> Some people in my old lab might have it and they check this mailing list.
> However I have also CCed Gabriel Landini (affiliated with Fiji) as he
> worked with me on this a bit and may have made a copy of the code.
>
> On Tue, Mar 18, 2014 at 5:04 PM, Carnë Draug <[hidden email]> wrote:
> > Hi Barry
> >
> > early last year you have contacted Antti Niemistö about his HistThresh
> > toolbox [1]. This was a Matlab toolbox for image thresholding which he
> > released under GPLv2+. In the mailing list, he agreed to release it
> > under BSD for ImageJ2 development.
> >
> > Unfortunately, it appears that he no longer works at the Tampere
> > University of Technology and the website for the project has been
> > taken down. His previous email address has also been cancelled and I
> > couldn't find his new contact. I tried to use wayback machine to
> > salvage his project, but the last backup [2] was before the last
> > release (he made one release prompted by your email which fixed a
> > couple of bugs in the minimum and maximum likelihood method). I have
> > made a github repository with the version I found [3]
> >
> > I made an Octave port which works fine for me [4] but I know some
> > Matlab users who would like to use it. Would you happen to have made a
> > backup of the last release he made?
> >
> > Thanks in advance,
> > Carnë
> >
> > [1] http://imagej.net/pipermail/imagej-devel/2013-February/001404.html
> > [2]
> > http://web.archive.org/web/20120628113512/http://www.cs.tut.fi/~ant/histth
> > resh/ [3] https://github.com/carandraug/histthresh
> > [4]
> > https://sourceforge.net/p/octave/image/ci/default/tree/inst/graythresh.m
Hi Barry, Carnë,
This is the only version I have of the maximum likelihood estimator
(th_maxlik.m) code from Anti's website. The date of the file in my computer is
2/Nov/2004.
His method uses the th_minimum.m procedure as a first approximation, which in
turn uses the bimodtest.m and others (also attached), but in my endeavours of
porting the th_maxlik method in Java for IJ I failed miserably in reproducing
the results that the code gave for matlab (shown in the pdf attached).
I spent some time with this particular method trying to resolve what was the
issue, but I never succeeded, so rather than having a broken method, I left it
our from the autothreshold plugin.

I hope it is useful.

Regards

Gabriel


_______________________________________________
ImageJ-devel mailing list
[hidden email]
http://imagej.net/mailman/listinfo/imagej-devel

th_maxlik.m (2K) Download Attachment
th_minimum.m (1K) Download Attachment
thresholds_README (2K) Download Attachment
ThreshComp.pdf (1M) Download Attachment
bimodtest.m (673 bytes) Download Attachment
A.m (443 bytes) Download Attachment
B.m (457 bytes) Download Attachment
C.m (460 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: last release of HistThresh

Carnë Draug
On 19 March 2014 16:02, Gabriel Landini <[hidden email]> wrote:

> On Tuesday 18 Mar 2014 18:31:58 Barry DeZonia wrote:
>> I am no longer working with the lab that worked with Antti. I have looked
>> at the backups I made when I left and I do not have a copy of Antti's work.
>> Some people in my old lab might have it and they check this mailing list.
>> However I have also CCed Gabriel Landini (affiliated with Fiji) as he
>> worked with me on this a bit and may have made a copy of the code.
>>
>> On Tue, Mar 18, 2014 at 5:04 PM, Carnë Draug <[hidden email]> wrote:
>> > Hi Barry
>> >
>> > early last year you have contacted Antti Niemistö about his HistThresh
>> > toolbox [1]. This was a Matlab toolbox for image thresholding which he
>> > released under GPLv2+. In the mailing list, he agreed to release it
>> > under BSD for ImageJ2 development.
>> >
>> > Unfortunately, it appears that he no longer works at the Tampere
>> > University of Technology and the website for the project has been
>> > taken down. His previous email address has also been cancelled and I
>> > couldn't find his new contact. I tried to use wayback machine to
>> > salvage his project, but the last backup [2] was before the last
>> > release (he made one release prompted by your email which fixed a
>> > couple of bugs in the minimum and maximum likelihood method). I have
>> > made a github repository with the version I found [3]
>> >
>> > I made an Octave port which works fine for me [4] but I know some
>> > Matlab users who would like to use it. Would you happen to have made a
>> > backup of the last release he made?
>> >
>> > Thanks in advance,
>> > Carnë
>> >
>> > [1] http://imagej.net/pipermail/imagej-devel/2013-February/001404.html
>> > [2]
>> > http://web.archive.org/web/20120628113512/http://www.cs.tut.fi/~ant/histth
>> > resh/ [3] https://github.com/carandraug/histthresh
>> > [4]
>> > https://sourceforge.net/p/octave/image/ci/default/tree/inst/graythresh.m
>
> Hi Barry, Carnë,
> This is the only version I have of the maximum likelihood estimator
> (th_maxlik.m) code from Anti's website. The date of the file in my computer is
> 2/Nov/2004.
> His method uses the th_minimum.m procedure as a first approximation, which in
> turn uses the bimodtest.m and others (also attached), but in my endeavours of
> porting the th_maxlik method in Java for IJ I failed miserably in reproducing
> the results that the code gave for matlab (shown in the pdf attached).
> I spent some time with this particular method trying to resolve what was the
> issue, but I never succeeded, so rather than having a broken method, I left it
> our from the autothreshold plugin.
>
> I hope it is useful.

Thank you all.

I have finally managed to contact the original author, Antti Niemistö,
who sent me all the files. For future reference, I changed the
repository mentioned before [3], using the exact files he sent me,
including both versions. I added a note to the README file about the
ImageJ ports.

Carnë


[3] https://github.com/carandraug/histthresh

_______________________________________________
ImageJ-devel mailing list
[hidden email]
http://imagej.net/mailman/listinfo/imagej-devel
Reply | Threaded
Open this post in threaded view
|

Re: last release of HistThresh

Gabriel Landini
On Saturday 05 Apr 2014 03:07:44 Carnë Draug wrote:
> I have finally managed to contact the original author, Antti Niemistö,
> who sent me all the files. For future reference, I changed the
> repository mentioned before [3], using the exact files he sent me,
> including both versions. I added a note to the README file about the
> ImageJ ports.

Just to clarify the wrong attribution, in the History section.
https://github.com/carandraug/histthresh/blob/master/README.md
It was me who originally ported Antti's methods to ImageJ 1.X, ending up in
the authothresolds plugins.

Not sure if Barry ported those from IJ1 to IJ2 or directly from Antti's code,
better confirm with him directly.

Cheers

Gabriel


_______________________________________________
ImageJ-devel mailing list
[hidden email]
http://imagej.net/mailman/listinfo/imagej-devel
Reply | Threaded
Open this post in threaded view
|

Re: last release of HistThresh

dscho
Administrator
Hi,

On Sat, 5 Apr 2014, Gabriel Landini wrote:

> On Saturday 05 Apr 2014 03:07:44 Carnë Draug wrote:
> > I have finally managed to contact the original author, Antti Niemistö,
> > who sent me all the files. For future reference, I changed the
> > repository mentioned before [3], using the exact files he sent me,
> > including both versions. I added a note to the README file about the
> > ImageJ ports.
>
> Just to clarify the wrong attribution, in the History section.
> https://github.com/carandraug/histthresh/blob/master/README.md
> It was me who originally ported Antti's methods to ImageJ 1.X, ending up in
> the authothresolds plugins.
>
> Not sure if Barry ported those from IJ1 to IJ2 or directly from Antti's
> code, better confirm with him directly.
Barry definitely used the Auto Threshold source code (with permission) to
base the BSD-licensed ImageJ2 version. Pull request here:

https://github.com/carandraug/histthresh/pull/1

Ciao,
Dscho


_______________________________________________
ImageJ-devel mailing list
[hidden email]
http://imagej.net/mailman/listinfo/imagej-devel