XY Scale in Histogram

[Sidnei Paciornik]

Dear colleagues,

When using histograms for segmentation it is many times relevant to expand
the x and y scales of the graph to reveal smaller peaks. This is very
common in mounted materials science samples in which there is a very large
peak for the tone of mounting polymer.

The question is:
Is there anyway to vary the Y (and possibly the X) scale in histogram,
especially for thresholding?

One workaround for the large peak mentioned above is to threshold it so
that it goes to zero (while preserving the remaining pixel intensities) and
then have an option in the histogram that ignores the zero value (skip
zero). Thus, the remaining peaks become more preeminent in the histogram.

Any suggestions?

Regards

Sidnei

Prof. Sidnei Paciornik
Grupo de Análise de Imagens e Microscopia Digital
DEMa <http://www.dema.puc-rio.br/> - Departamento de Engenharia de Materiais
PUC-Rio <http://www.puc-rio.br/>
Rua Marquês de São Vicente 225
Prédio Leme, Sala 501L
Gávea - Rio de Janeiro - RJ
22451-900 - Brasil
tel: [phone]

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

[James Ewing]

Sidnei - Do you have an a priori idea about the shape of the major peak?  If so, you could fit the peak and subtract it from the data, then examine the remainder.  
 - Jim


On Jun 23, 2013, at 10:26 PM, Sidnei Paciornik wrote:

> Dear colleagues,
>
> When using histograms for segmentation it is many times relevant to expand
> the x and y scales of the graph to reveal smaller peaks. This is very
> common in mounted materials science samples in which there is a very large
> peak for the tone of mounting polymer.
>
> The question is:
> Is there anyway to vary the Y (and possibly the X) scale in histogram,
> especially for thresholding?
>
> One workaround for the large peak mentioned above is to threshold it so
> that it goes to zero (while preserving the remaining pixel intensities) and
> then have an option in the histogram that ignores the zero value (skip
> zero). Thus, the remaining peaks become more preeminent in the histogram.
>
> Any suggestions?
>
> Regards
>
> Sidnei
>
> Prof. Sidnei Paciornik
> Grupo de Análise de Imagens e Microscopia Digital
> DEMa <http://www.dema.puc-rio.br/> - Departamento de Engenharia de Materiais
> PUC-Rio <http://www.puc-rio.br/>
> Rua Marquês de São Vicente 225
> Prédio Leme, Sala 501L
> Gávea - Rio de Janeiro - RJ
> 22451-900 - Brasil
> tel: [phone]
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

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

[Sidnei Paciornik]

Hi James,

Actually not. The peak can have any shape and, for that matter, I would
need to eliminate it in the histogram preferably during the threshold
operation. So it would have to be a simple interface.

I think the best idea, if possible to implement, it to have at least the y
scale of the histogram variable and controlled by the user. Certain
programs (KS400, SIS) offer this option.

Thank you,

Sidnei


On Mon, Jun 24, 2013 at 6:31 PM, James Ewing <[hidden email]>
 wrote:

> Sidnei - Do you have an a priori idea about the shape of the major peak?
>  If so, you could fit the peak and subtract it from the data, then examine
> the remainder.
>  - Jim
>
>
> On Jun 23, 2013, at 10:26 PM, Sidnei Paciornik wrote:
>
> > Dear colleagues,
> >
> > When using histograms for segmentation it is many times relevant to
> expand
> > the x and y scales of the graph to reveal smaller peaks. This is very
> > common in mounted materials science samples in which there is a very
> large
> > peak for the tone of mounting polymer.
> >
> > The question is:
> > Is there anyway to vary the Y (and possibly the X) scale in histogram,
> > especially for thresholding?
> >
> > One workaround for the large peak mentioned above is to threshold it so
> > that it goes to zero (while preserving the remaining pixel intensities)
> and
> > then have an option in the histogram that ignores the zero value (skip
> > zero). Thus, the remaining peaks become more preeminent in the histogram.
> >
> > Any suggestions?
> >
> > Regards
> >
> > Sidnei
> >
> > Prof. Sidnei Paciornik
> > Grupo de Análise de Imagens e Microscopia Digital
> > DEMa <http://www.dema.puc-rio.br/> - Departamento de Engenharia de
> Materiais
> > PUC-Rio <http://www.puc-rio.br/>
> > Rua Marquês de São Vicente 225
> > Prédio Leme, Sala 501L
> > Gávea - Rio de Janeiro - RJ
> > 22451-900 - Brasil
> > tel: [phone]
> >
> > --
> > ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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

[dqvn]

Dear all,

I'm facing this problem: I have an image as below, having a rectangular
in center of the image, which can be rotated, scaled, and so on; through
the ImageJ I want to keep record of the coordinates of the 4 corners of
the rectangle in image such as [TL(x1,y1); TR(x2,y2); BL(x3,y3);
BR(x4,y4)]. After that, these 4 points will help me solving my problem
so far.



In my aspect, the ImageJ could let me find the center point  and draw
the cover of the object (the rectangle). But I need the axis of 4
corners at all. Because I am a newbie in ImageJ, thus could you please
let me know steps that I need to prepare for solving my requirement?.

Example of image:



Thanks and best regards,
  Quang Nguyen.

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

[Gabriel Landini]

On Monday 24 Jun 2013 23:49:40 Sidnei Paciornik wrote:
> Actually not. The peak can have any shape and, for that matter, I would
> need to eliminate it in the histogram preferably during the threshold
> operation. So it would have to be a simple interface.
>
> I think the best idea, if possible to implement, it to have at least the y
> scale of the histogram variable and controlled by the user. Certain
> programs (KS400, SIS) offer this option.

Currently you can't do this with of the automated thresholding methods, BUT
the original Auto_threshold plugin downloaded from my page:
http://www.dentistry.bham.ac.uk/landinig/software/software.html
(not the built in on which is a derivative of it) it can be set to ignore the
white (255) and/or black (0) bins by setting their frequencies to 0 in the
computation.
So maybe you can just preserve the image section you want to threshold and set
the rest to 0, click "Ignore black" and that should do the trick.
The other advantage is that it can threshold 16 bit image histograms (instead
of reducing them to 8 bits first).

Cheers
Gabriel

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