Cell automatic selection on bad image

[Василий Попков]

Good time of a day, friends and colleagues.
Anyone can give me some advices how (and if - maybe it will be wiser not to
waste time) I can automaticly select cells (for roi manager) on such image?
For "eyes" it dosent look so bad, though yes, phase is awful, thankfully
fluorescent channel is ok, but it's hard to distinguish cells there. But I
have tried all more or less standart methods and they all were doing quite
bad selection.
Or, to be more cpecific, they select almost nothing - I'll be more or less
ok with "bad" selection if it will be bad selection of cells.
Otherwise I am going to manually select them with "cell magic wand tool".
Its just a bit "not clever" work.

Thanks,
Vasily

--
Vasily A Popkov
Lab.Structure and function of mitochondria
A.N.Belozersky Institute of Physico-Chemical Biology
Moscow State University

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

[ctrueden]

Hi Vasily,

Did you try the Trainable Weka Segmentation?

http://imagej.net/Trainable_Weka_Segmentation
http://forum.imagej.net/t/analysis-of-aggregates/303/3?u=ctrueden
http://imagej.net/Segmentation

Note the following helpful comment in that forum thread: "In the settings,
I had added the Variance and Structure features, increased the maximum
sigma to 32, and checked Homogenize classes"

I did not have time right now to try it with your data, but perhaps it
helps you.

Regards,
Curtis

--
Curtis Rueden
LOCI software architect - http://loci.wisc.edu/software
ImageJ2 lead, Fiji maintainer - http://imagej.net/User:Rueden
Did you know ImageJ has a forum? http://forum.imagej.net/


On Mon, Oct 10, 2016 at 7:38 AM, Василий Попков <[hidden email]>
wrote:

> Good time of a day, friends and colleagues.
> Anyone can give me some advices how (and if - maybe it will be wiser not to
> waste time) I can automaticly select cells (for roi manager) on such image?
> For "eyes" it dosent look so bad, though yes, phase is awful, thankfully
> fluorescent channel is ok, but it's hard to distinguish cells there. But I
> have tried all more or less standart methods and they all were doing quite
> bad selection.
> Or, to be more cpecific, they select almost nothing - I'll be more or less
> ok with "bad" selection if it will be bad selection of cells.
> Otherwise I am going to manually select them with "cell magic wand tool".
> Its just a bit "not clever" work.
>
> Thanks,
> Vasily
>
> --
> Vasily A Popkov
> Lab.Structure and function of mitochondria
> A.N.Belozersky Institute of Physico-Chemical Biology
> Moscow State University
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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

[Василий Попков]

Thank you for answer!
Yes, I have tried it, and it didn't work well. Though I have little
experience with weka. So, since you are advising it, I'll check some
manuals and will try more. Without some expert opinion it is hard to
understand was it my mistake or not suitable method.

V

2016-10-10 22:24 GMT+03:00 Curtis Rueden <[hidden email]>:

> Hi Vasily,
>
> Did you try the Trainable Weka Segmentation?
>
> http://imagej.net/Trainable_Weka_Segmentation
> http://forum.imagej.net/t/analysis-of-aggregates/303/3?u=ctrueden
> http://imagej.net/Segmentation
>
> Note the following helpful comment in that forum thread: "In the settings,
> I had added the Variance and Structure features, increased the maximum
> sigma to 32, and checked Homogenize classes"
>
> I did not have time right now to try it with your data, but perhaps it
> helps you.
>
> Regards,
> Curtis
>
> --
> Curtis Rueden
> LOCI software architect - http://loci.wisc.edu/software
> ImageJ2 lead, Fiji maintainer - http://imagej.net/User:Rueden
> Did you know ImageJ has a forum? http://forum.imagej.net/
>
>
> On Mon, Oct 10, 2016 at 7:38 AM, Василий Попков <[hidden email]>
> wrote:
>
> > Good time of a day, friends and colleagues.
> > Anyone can give me some advices how (and if - maybe it will be wiser not
> to
> > waste time) I can automaticly select cells (for roi manager) on such
> image?
> > For "eyes" it dosent look so bad, though yes, phase is awful, thankfully
> > fluorescent channel is ok, but it's hard to distinguish cells there. But
> I
> > have tried all more or less standart methods and they all were doing
> quite
> > bad selection.
> > Or, to be more cpecific, they select almost nothing - I'll be more or
> less
> > ok with "bad" selection if it will be bad selection of cells.
> > Otherwise I am going to manually select them with "cell magic wand tool".
> > Its just a bit "not clever" work.
> >
> > Thanks,
> > Vasily
> >
> > --
> > Vasily A Popkov
> > Lab.Structure and function of mitochondria
> > A.N.Belozersky Institute of Physico-Chemical Biology
> > Moscow State University
> >
> > --
> > ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> >
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
Vasily A Popkov
Lab.Structure and function of mitochondria
A.N.Belozersky Institute of Physico-Chemical Biology
Moscow State University

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