Removing Outliers With Increased Accuracy

[MikeZ]

Hi there!

I've been working with images of nanoparticle dispersions on different substrates, and I'm having a particularly hard time with one of my samples.  The problem is, after having removed what was a very detailed and irregular substrate, I have a significant amount of noise left over in my binary image.  The noise is pretty much made entirely of 1-3 pixel outliers.  I've tried using the Remove Outliers tool within Noise, but it isn't accurate enough - the lowest threshold for the process causes some of my smaller particles to be lost.  Is there any way I can more accurately describe the size of these outliers so that I can remove them, while leaving my particles untouched?  Are there any alternate processes you might recommend?  I've tried thresholding, but unfortunately the substrate is detailed enough that it's essentially useless (in order to keep all of the particles, I have to keep a significant portion of the substrate).  

Thanks!

Note:  My image is 8-bit

[Christian Tischer]

Hi Mike,

not sure i 100% understand, but what about using Analyse particles and
rejecting everything which has a size smaller than 3 pixels?

Tischi


On Mon, Jul 16, 2012 at 4:44 PM, MikeZ <[hidden email]> wrote:

> Hi there!
>
> I've been working with images of nanoparticle dispersions on different
> substrates, and I'm having a particularly hard time with one of my samples.
> The problem is, after having removed what was a *very* detailed and
> irregular substrate, I have a significant amount of noise left over in my
> binary image.  The noise is pretty much made entirely of 1-3 pixel
> outliers.
> I've tried using the Remove Outliers tool within Noise, but it isn't
> accurate enough - the lowest threshold for the process causes some of my
> smaller particles to be lost.  Is there any way I can more accurately
> describe the size of these outliers so that I can remove them, while
> leaving
> my particles untouched?  Are there any alternate processes you might
> recommend?  I've tried thresholding, but unfortunately the substrate is
> detailed enough that it's essentially useless (in order to keep all of the
> particles, I have to keep a significant portion of the substrate).
>
> Thanks!
>
> --
> View this message in context:
> http://imagej.1557.n6.nabble.com/Removing-Outliers-With-Increased-Accuracy-tp4999435.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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

[BROUILLY NICOLAS]

Hi Mike,

It would be much easier to get an idea of your problem with a representative image.

Indeed, try with the analyse particles tool, playing with the settings...

Nico


Le 16 juil. 2012 à 17:51, Christian Tischer a écrit :

Hi Mike,

not sure i 100% understand, but what about using Analyse particles and
rejecting everything which has a size smaller than 3 pixels?

Tischi


On Mon, Jul 16, 2012 at 4:44 PM, MikeZ <[hidden email]<mailto:[hidden email]>> wrote:

Hi there!

I've been working with images of nanoparticle dispersions on different
substrates, and I'm having a particularly hard time with one of my samples.
The problem is, after having removed what was a *very* detailed and
irregular substrate, I have a significant amount of noise left over in my
binary image.  The noise is pretty much made entirely of 1-3 pixel
outliers.
I've tried using the Remove Outliers tool within Noise, but it isn't
accurate enough - the lowest threshold for the process causes some of my
smaller particles to be lost.  Is there any way I can more accurately
describe the size of these outliers so that I can remove them, while
leaving
my particles untouched?  Are there any alternate processes you might
recommend?  I've tried thresholding, but unfortunately the substrate is
detailed enough that it's essentially useless (in order to keep all of the
particles, I have to keep a significant portion of the substrate).

Thanks!

--
View this message in context:
http://imagej.1557.n6.nabble.com/Removing-Outliers-With-Increased-Accuracy-tp4999435.html
Sent from the ImageJ mailing list archive at Nabble.com<http://Nabble.com>.

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


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

Nicolas Brouilly, PhD Student
Equipe Gieseler, "Pathologies du muscle chez C. elegans"

Centre de Génétique et de Physiologie Moléculaires et Cellulaires
UMR CNRS 5534
Université Claude Bernard Lyon 1
Bâtiment Gregor Mendel, 5ème étage
16, rue Raphaël Dubois
F-69622 Villeurbanne Cedex

[hidden email]<mailto:[hidden email]>
Lab : +0033 (0)4 72 43 29 48
Cel : +0033 (0)6 42 91 53 19


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

[MikeZ]

@Tischi

Is there a way to delete particles below a certain threshold value?  If I could delete all of the particles that are less than or equal to 3 units^2 in area, I think that would provide a suitable image.  I realize I can do that in Analyze Particles, but I'd like to make sure by overlaying the new particles over the original image (which, as far as I know, can't be done with Analyze Particles).  I've tried "Remove Outliers", but I didn't seem to do the trick for this, but I may have entered the parameters incorrectly... If there's a more effective way to do this, though, I would very much appreciate any advice

[Rodrigo Gonçalves-4]

Hi, did you try with particle remover?

http://rsbweb.nih.gov/ij/plugins/particle-remover.html

 
________________________
Rodrigo J. Gonçalves

________________________

>________________________________
> De: MikeZ <[hidden email]>
>Para: [hidden email]
>Enviado: Lunes, 16 de julio, 2012 9:53 P.M.
>Asunto: Re: Removing Outliers With Increased Accuracy
>
>@Tischi
>
>Is there a way to delete particles below a certain threshold value?  If I
>could delete all of the particles that are less than or equal to 3 units^2
>in area, I /think/ that would provide a suitable image.  I realize I can do
>that in Analyze Particles, but I'd like to make sure by overlaying the new
>particles over the original image (which, as far as I know, can't be done
>with Analyze Particles).  I've tried "Remove Outliers", but I didn't seem to
>do the trick for this, but I may have done something wrong...
>
>--
>View this message in context: http://imagej.1557.n6.nabble.com/Removing-Outliers-With-Increased-Accuracy-tp4999435p4999444.html
>Sent from the ImageJ mailing list archive at Nabble.com.
>
>--
>ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
>
>

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

[Robert Baer]

On 7/16/2012 2:53 PM, MikeZ wrote:

> @Tischi
>
> Is there a way to delete particles below a certain threshold value?  If I
> could delete all of the particles that are less than or equal to 3 units^2
> in area, I /think/ that would provide a suitable image.  I realize I can do
> that in Analyze Particles, but I'd like to make sure by overlaying the new
> particles over the original image (which, as far as I know, can't be done
> with Analyze Particles).  I've tried "Remove Outliers", but I didn't seem to
> do the trick for this, but I may have done something wrong...
>
> --
> View this message in context: http://imagej.1557.n6.nabble.com/Removing-Outliers-With-Increased-Accuracy-tp4999435p4999444.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

It seems to me you were on the right track with analyze particles unless
I misuderstood your goal.  For example, open the blobs sample and
perform Process | Binary | Make binary to make it a good example.  Now
do Analyze | analyze particles where you set the upper limit for area to
say, 500 units, and you set the show field to Mask.  Isn't the image
produced just the type of altered image you need albeit getting rid of
the bigger circles instead of the smaller ones with the particular way I
formulated the example?

Hope this helps?

Rob

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

[Christian Tischer]

Hi Mike,

in Analyse Particles select: show...Overlay Outlines.

Tischi

On Mon, Jul 16, 2012 at 9:53 PM, MikeZ <[hidden email]> wrote:

> @Tischi
>
> Is there a way to delete particles below a certain threshold value?  If I
> could delete all of the particles that are less than or equal to 3 units^2
> in area, I /think/ that would provide a suitable image.  I realize I can do
> that in Analyze Particles, but I'd like to make sure by overlaying the new
> particles over the original image (which, as far as I know, can't be done
> with Analyze Particles).  I've tried "Remove Outliers", but I didn't seem
> to
> do the trick for this, but I may have done something wrong...
>
> --
> View this message in context:
> http://imagej.1557.n6.nabble.com/Removing-Outliers-With-Increased-Accuracy-tp4999435p4999444.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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

[MikeZ]

This solved my issue, thank you very much, everyone :) !

- Mike