individual image pixel count in a time series image

[srai]

Hello everyone

I am looking at a time series image. After thresholding to make binary image, the images have pixel value of either 255 or 0. I just want the pixel value count of either one, for each individual image in the stack to look at how the pixel count is changing.

I would be grateful if anyone had any suggestions.

Srai

[Daniel Kalthoff]

Hi Srai,

I'm not sure if I got you right, but I'd suggest this:

1. Select the whole image ("Select All")
2. Plot z-axis profile
3. Divide the resulting profile by (width*height / 255) --> gives you the number of 255 voxels

Problem solved?


Am 10.05.2011 um 15:36 schrieb lookingatwaves:

> Hello everyone
>
> I am looking at a time series image. After thresholding to make binary
> image, the images have pixel value of either 255 or 0. I just want the pixel
> value count of either one, for each individual image in the stack to look at
> how the pixel count is changing.
>
> I would be grateful if anyone had any suggestions.
>
> Srai
>
> --
> View this message in context: http://imagej.588099.n2.nabble.com/individual-image-pixel-count-in-a-time-series-image-tp6347840p6347840.html
> Sent from the ImageJ mailing list archive at Nabble.com.

[Julien Colombelli]

Hi Srai,

My way to measure this would be:   (remember that the plot Z-Axis profile
will measure only what you select in 'Set Measurements...' )

-  tick the *Integrated density* option in 'Set Measurements...',
- Select All, or unselect current selection.
-  run 'Plot Z-axis Profile'  on the whole stack.
- The column 'Int Dens' will include the results you want, but you will have
to divide by 255 to get the '255 pixels' counts (integrated density sums all
pixels values),
to get the 'Zero pixels' count you can either subtract the resulting values
to the total number of pixels,
or invert your image and run 'Plot Z-Axis Profile' again,

Remember to save the results table before running 'Plot Z-axis profile'
again.
And be careful, the plot legend might not reflect the real values which are
in the results table (at least not in my case).

You can divide the image by 255 (in menu 'Process/Maths') before counting to
get the right number directly. If you do it for the Zero pixel, you must
invert first, then divide.

Hope this helps,

J


On 10 May 2011 15:36, lookingatwaves <[hidden email]> wrote:

> Hello everyone
>
> I am looking at a time series image. After thresholding to make binary
> image, the images have pixel value of either 255 or 0. I just want the
> pixel
> value count of either one, for each individual image in the stack to look
> at
> how the pixel count is changing.
>
> I would be grateful if anyone had any suggestions.
>
> Srai
>
> --
> View this message in context:
> http://imagej.588099.n2.nabble.com/individual-image-pixel-count-in-a-time-series-image-tp6347840p6347840.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>

--
Julien Colombelli
Advanced Digital Microscopy
Institute for Research in Biomedicine - IRB Barcelona
Baldiri Reixac, 10
Tel:  [phone]
Fax: [phone]
E-08028 Barcelona - Spain
[hidden email]

[srai]

Hello Daniel

Thanks for the help. It works:)
Srai

> Date: Tue, 10 May 2011 16:42:05 +0200
> From: [hidden email]
> Subject: Re: individual image pixel count in a time series image
> To: [hidden email]
>
> Hi Srai,
>
> I'm not sure if I got you right, but I'd suggest this:
>
> 1. Select the whole image ("Select All")
> 2. Plot z-axis profile
> 3. Divide the resulting profile by (width*height / 255) --> gives you the number of 255 voxels
>
> Problem solved?
>
>
> Am 10.05.2011 um 15:36 schrieb lookingatwaves:
>
> > Hello everyone
> >
> > I am looking at a time series image. After thresholding to make binary
> > image, the images have pixel value of either 255 or 0. I just want the pixel
> > value count of either one, for each individual image in the stack to look at
> > how the pixel count is changing.
> >
> > I would be grateful if anyone had any suggestions.
> >
> > Srai
> >
> > --
> > View this message in context: http://imagej.588099.n2.nabble.com/individual-image-pixel-count-in-a-time-series-image-tp6347840p6347840.html
> > Sent from the ImageJ mailing list archive at Nabble.com.

     

[chai_san]

Hi,
the easiest should be to set measurements to "Area, Mean grey value", measure (control+K, apple+K depending on your system). The output you get should then by simple arithmetic give you your no. of white pixels (multiply the mean with the total pixel number=area).
Cheers,
Chaitanya

[srai]

Hello Julien

Thanks for the detailed description. I managed to get it work:)
Srai

> Date: Tue, 10 May 2011 16:58:45 +0200
> From: [hidden email]
> Subject: Re: individual image pixel count in a time series image
> To: [hidden email]
>
> Hi Srai,
>
> My way to measure this would be:   (remember that the plot Z-Axis profile
> will measure only what you select in 'Set Measurements...' )
>
> -  tick the *Integrated density* option in 'Set Measurements...',
> - Select All, or unselect current selection.
> -  run 'Plot Z-axis Profile'  on the whole stack.
> - The column 'Int Dens' will include the results you want, but you will have
> to divide by 255 to get the '255 pixels' counts (integrated density sums all
> pixels values),
> to get the 'Zero pixels' count you can either subtract the resulting values
> to the total number of pixels,
> or invert your image and run 'Plot Z-Axis Profile' again,
>
> Remember to save the results table before running 'Plot Z-axis profile'
> again.
> And be careful, the plot legend might not reflect the real values which are
> in the results table (at least not in my case).
>
> You can divide the image by 255 (in menu 'Process/Maths') before counting to
> get the right number directly. If you do it for the Zero pixel, you must
> invert first, then divide.
>
> Hope this helps,
>
> J
>
>
> On 10 May 2011 15:36, lookingatwaves <[hidden email]> wrote:
>
> > Hello everyone
> >
> > I am looking at a time series image. After thresholding to make binary
> > image, the images have pixel value of either 255 or 0. I just want the
> > pixel
> > value count of either one, for each individual image in the stack to look
> > at
> > how the pixel count is changing.
> >
> > I would be grateful if anyone had any suggestions.
> >
> > Srai
> >
> > --
> > View this message in context:
> > http://imagej.588099.n2.nabble.com/individual-image-pixel-count-in-a-time-series-image-tp6347840p6347840.html
> > Sent from the ImageJ mailing list archive at Nabble.com.
> >
>
>
>
> --
> Julien Colombelli
> Advanced Digital Microscopy
> Institute for Research in Biomedicine - IRB Barcelona
> Baldiri Reixac, 10
> Tel:  [phone]
> Fax: [phone]
> E-08028 Barcelona - Spain
> [hidden email]