ImageJ

16 bit image visualization

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

8 messages Options

Knecht, David

16 bit image visualization

When we collect 16 bit images, it is sometimes because we want the dynamic range to visualize structures or cells that are of wide ranging intensities. However, as far as I know, it is difficult to isolate different intensity ranges in a 16 bit image and then play with the brightness/contrast of those pixels. What I think I am looking for is a tool like the Adjust Contrast/Brightness sliders, but be able to set values above maximum the slider value to 0. Or a tool like the Adjust Threshold tool, but instead of highlighting the pixels it would set pixels outside the selected range to black. Another way of looking at this would be the equivalent of the min and max transparency bound in a projection. I am wondering if such a tool is available in ImageJ or somewhere in the plugin universe and I have just not found it. Thanks- Dave

Dr. David Knecht
Professor , Department of Molecular and Cell Biology
University of Connecticut
91 N. Eagleville Rd.
U-3125
Storrs, CT 06269-3125
860-486-2200

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

Jeremy Adler-2

Re: 16 bit image visualization

It is very easy to set the display of intensities above the selected threshold to zero - or any other value.
Use Image/ Color/ Edit LUT
and set the display values for the maximum intensity to 0,
then save the LUT.

________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Knecht, David [[hidden email]]
Sent: 27 April 2017 23:14
To: [hidden email]
Subject: 16 bit image visualization

When we collect 16 bit images, it is sometimes because we want the dynamic range to visualize structures or cells that are of wide ranging intensities. However, as far as I know, it is difficult to isolate different intensity ranges in a 16 bit image and then play with the brightness/contrast of those pixels. What I think I am looking for is a tool like the Adjust Contrast/Brightness sliders, but be able to set values above maximum the slider value to 0. Or a tool like the Adjust Threshold tool, but instead of highlighting the pixels it would set pixels outside the selected range to black. Another way of looking at this would be the equivalent of the min and max transparency bound in a projection. I am wondering if such a tool is available in ImageJ or somewhere in the plugin universe and I have just not found it. Thanks- Dave

Dr. David Knecht
Professor , Department of Molecular and Cell Biology
University of Connecticut
91 N. Eagleville Rd.
U-3125
Storrs, CT 06269-3125
860-486-2200

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

Jeremy Adler-2

Re: 16 bit image visualization

clarification of earlier post

Adjust brightness and contrast works with the LUT used by your image.

You need to save your new lut (max set to zero) in the LUTs folder of ImageJ and when you restart it will appear in the list of LUTs.

So you need to display your image in the new LUT before running brightness & constrast.

________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Jeremy Adler [[hidden email]]
Sent: 28 April 2017 05:51
To: [hidden email]
Subject: Re: 16 bit image visualization

It is very easy to set the display of intensities above the selected threshold to zero - or any other value.
Use Image/ Color/ Edit LUT
and set the display values for the maximum intensity to 0,
then save the LUT.

________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Knecht, David [[hidden email]]
Sent: 27 April 2017 23:14
To: [hidden email]
Subject: 16 bit image visualization

When we collect 16 bit images, it is sometimes because we want the dynamic range to visualize structures or cells that are of wide ranging intensities. However, as far as I know, it is difficult to isolate different intensity ranges in a 16 bit image and then play with the brightness/contrast of those pixels. What I think I am looking for is a tool like the Adjust Contrast/Brightness sliders, but be able to set values above maximum the slider value to 0. Or a tool like the Adjust Threshold tool, but instead of highlighting the pixels it would set pixels outside the selected range to black. Another way of looking at this would be the equivalent of the min and max transparency bound in a projection. I am wondering if such a tool is available in ImageJ or somewhere in the plugin universe and I have just not found it. Thanks- Dave

Dr. David Knecht
Professor , Department of Molecular and Cell Biology
University of Connecticut
91 N. Eagleville Rd.
U-3125
Storrs, CT 06269-3125
860-486-2200

--
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

Knecht, David

Re: 16 bit image visualization

When I go in to Image/Color/edit LUT, I only get 8 bits worth of LUT to adjust even when it is a 16 bit image. Am I missing something? Dave

Dr. David Knecht
Professor , Department of Molecular and Cell Biology
University of Connecticut
91 N. Eagleville Rd.
U-3125
Storrs, CT 06269-3125
860-486-2200

> On Apr 28, 2017, at 5:09 AM, Jeremy Adler <[hidden email]> wrote:
>
> clarification of earlier post
>
> Adjust brightness and contrast works with the LUT used by your image.
>
> You need to save your new lut (max set to zero) in the LUTs folder of ImageJ and when you restart it will appear in the list of LUTs.
>
> So you need to display your image in the new LUT before running brightness & constrast.
>
>
>
>
> ________________________________________
> From: ImageJ Interest Group [[hidden email]] on behalf of Jeremy Adler [[hidden email]]
> Sent: 28 April 2017 05:51
> To: [hidden email]
> Subject: Re: 16 bit image visualization
>
> It is very easy to set the display of intensities above the selected threshold to zero - or any other value.
> Use Image/ Color/ Edit LUT
> and set the display values for the maximum intensity to 0,
> then save the LUT.
>
>
> ________________________________________
> From: ImageJ Interest Group [[hidden email]] on behalf of Knecht, David [[hidden email]]
> Sent: 27 April 2017 23:14
> To: [hidden email]
> Subject: 16 bit image visualization
>
> When we collect 16 bit images, it is sometimes because we want the dynamic range to visualize structures or cells that are of wide ranging intensities. However, as far as I know, it is difficult to isolate different intensity ranges in a 16 bit image and then play with the brightness/contrast of those pixels. What I think I am looking for is a tool like the Adjust Contrast/Brightness sliders, but be able to set values above maximum the slider value to 0. Or a tool like the Adjust Threshold tool, but instead of highlighting the pixels it would set pixels outside the selected range to black. Another way of looking at this would be the equivalent of the min and max transparency bound in a projection. I am wondering if such a tool is available in ImageJ or somewhere in the plugin universe and I have just not found it. Thanks- Dave
>
> Dr. David Knecht
> Professor , Department of Molecular and Cell Biology
> University of Connecticut
> 91 N. Eagleville Rd.
> U-3125
> Storrs, CT 06269-3125
> 860-486-2200
>
>
> --
> 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

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

Jeremy Adler-2

Re: 16 bit image visualization

The limitation is the human visual system - an 8 bit range is more than anyone can identify in an image displayed on a monitor.

Realistically we cannot accurately discriminate 24 intensity levels on a monitor, except under highly artifical conditions - our visual system has not evolved for accurate intensity resolution: edges, objects, shapes we can manage under very uneven lighting but as a consequence we are very poor at intensities.

Software can work with the whole 16 bit range but we cannot, and we are blind to much of the information present images and need tools like Brightness and Contrast and histograms.
False colour look up tables help when intensity is the primary consideration.

Jeremy Adler
Uppsala U.

________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Knecht, David [[hidden email]]
Sent: 01 May 2017 20:06
To: [hidden email]
Subject: Re: 16 bit image visualization

When I go in to Image/Color/edit LUT, I only get 8 bits worth of LUT to adjust even when it is a 16 bit image. Am I missing something? Dave

Dr. David Knecht
Professor , Department of Molecular and Cell Biology
University of Connecticut
91 N. Eagleville Rd.
U-3125
Storrs, CT 06269-3125
860-486-2200

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

Knecht, David

Re: 16 bit image visualization

Hi Jeremy- I realize that you can only see 8 bits, but you should be able to visualize 8 bits worth of data from different ranges of the 65000 gray levels in the image. I can easily look at 8 bits of bright pixels, but I can’t look at 8 bits of medium brightness or 8 bits of just over background information. The dim stuff is overwhelmed by the saturated bright pixels if you simply contrast the image to bring up those values, even though they are plenty bright relative to background.

Dr. David Knecht
Professor , Department of Molecular and Cell Biology
University of Connecticut
91 N. Eagleville Rd.
U-3125
Storrs, CT 06269-3125
860-486-2200

> On May 1, 2017, at 3:49 PM, Jeremy Adler <[hidden email]> wrote:
>
> The limitation is the human visual system - an 8 bit range is more than anyone can identify in an image displayed on a monitor.
>
> Realistically we cannot accurately discriminate 24 intensity levels on a monitor, except under highly artifical conditions - our visual system has not evolved for accurate intensity resolution: edges, objects, shapes we can manage under very uneven lighting but as a consequence we are very poor at intensities.
>
> Software can work with the whole 16 bit range but we cannot, and we are blind to much of the information present images and need tools like Brightness and Contrast and histograms.
> False colour look up tables help when intensity is the primary consideration.
>
> Jeremy Adler
> Uppsala U.
>
>
>
>
> ________________________________________
> From: ImageJ Interest Group [[hidden email]] on behalf of Knecht, David [[hidden email]]
> Sent: 01 May 2017 20:06
> To: [hidden email]
> Subject: Re: 16 bit image visualization
>
> When I go in to Image/Color/edit LUT, I only get 8 bits worth of LUT to adjust even when it is a 16 bit image. Am I missing something? Dave
>
> Dr. David Knecht
> Professor , Department of Molecular and Cell Biology
> University of Connecticut
> 91 N. Eagleville Rd.
> U-3125
> Storrs, CT 06269-3125
> 860-486-2200
>
>> On Apr 28, 2017, at 5:09 AM, Jeremy Adler <[hidden email]> wrote:
>>
>> clarification of earlier post
>>
>> Adjust brightness and contrast works with the LUT used by your image.
>>
>> You need to save your new lut (max set to zero) in the LUTs folder of ImageJ and when you restart it will appear in the list of LUTs.
>>
>> So you need to display your image in the new LUT before running brightness & constrast.
>>
>>
>>
>>
>> ________________________________________
>> From: ImageJ Interest Group [[hidden email]] on behalf of Jeremy Adler [[hidden email]]
>> Sent: 28 April 2017 05:51
>> To: [hidden email]
>> Subject: Re: 16 bit image visualization
>>
>> It is very easy to set the display of intensities above the selected threshold to zero - or any other value.
>> Use Image/ Color/ Edit LUT
>> and set the display values for the maximum intensity to 0,
>> then save the LUT.
>>
>>
>> ________________________________________
>> From: ImageJ Interest Group [[hidden email]] on behalf of Knecht, David [[hidden email]]
>> Sent: 27 April 2017 23:14
>> To: [hidden email]
>> Subject: 16 bit image visualization
>>
>> When we collect 16 bit images, it is sometimes because we want the dynamic range to visualize structures or cells that are of wide ranging intensities. However, as far as I know, it is difficult to isolate different intensity ranges in a 16 bit image and then play with the brightness/contrast of those pixels. What I think I am looking for is a tool like the Adjust Contrast/Brightness sliders, but be able to set values above maximum the slider value to 0. Or a tool like the Adjust Threshold tool, but instead of highlighting the pixels it would set pixels outside the selected range to black. Another way of looking at this would be the equivalent of the min and max transparency bound in a projection. I am wondering if such a tool is available in ImageJ or somewhere in the plugin universe and I have just not found it. Thanks- Dave
>>
>> Dr. David Knecht
>> Professor , Department of Molecular and Cell Biology
>> University of Connecticut
>> 91 N. Eagleville Rd.
>> U-3125
>> Storrs, CT 06269-3125
>> 860-486-2200
>>
>>
>> --
>> 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
>
> --
> 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

Gabriel Landini

Re: 16 bit image visualization

On Monday, 1 May 2017 21:07:26 BST you wrote:
> Hi Jeremy- I realize that you can only see 8 bits, but you should be able to
> visualize 8 bits worth of data from different ranges of the 65000 gray
> levels in the image. I can easily look at 8 bits of bright pixels, but I
> can’t look at 8 bits of medium brightness or 8 bits of just over background
> information.

If you want multiple ranges, then the LUT is not linear anymore, so perhaps
something like histogram equalisation or a log transform of the image would
help? Then convert to 8 bit.

On 8 bit level of perception, I understand that some people can see more. High
end radiological displays show more than 8bit of greyscale levels (e.g. 10 or
12 bits).
When dealing with colour LUTs you can easily get into noticeable stepping
effects on smooth gradients too (depending on the range of colours the LUT is
representing).

Cheers
Gabriel

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

Jeremy Adler-2

Re: 16 bit image visualization

In reply to this post by Knecht, David

________________________________________
From: Jeremy Adler
Sent: 02 May 2017 04:47
To: [hidden email]
Subject: RE: 16 bit image visualization

Hi David,
The overwhelming effect of the saturated pixels outside the intensity range selected using Brightness and Contrast can be removed.
As in mentioned in my first post - create a LUT where the peak intensity is set to zero, then all the saturated pixels are displayed in black.
As an illustration create a new LUT from a greyscale LUT (use Image\ Color \ Edit LUT) that displays the maximum as dark green and minimum in dark blue - save it into the LUTs folder and restart ImageJ. Try the LUT with an image that is is a 16 bit ramp (file new image, and as wide as your screen will support), select the LUT and play with Brightness & Contrast. The chosen display range uses 254 grey intensities, zero appears in a low intenisity blue and maximum in green - this LUT makes it apparent which pixels fall outside the range selected and is better than simply setting the max and min to zero.

Gabriel points out that there are high end monitors that can provide a wider display range, but illusions like Adelson's Checkers demonstrate that even though a monitor displays intensities correctly and precisely our visual system is too responsive to local contrast to correctly identify intensities. A good false colour LUT corrects the illusion generated by Adelson's Checkers.

Jeremy

________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Knecht, David [[hidden email]]
Sent: 01 May 2017 23:07
To: [hidden email]
Subject: Re: 16 bit image visualization

Hi Jeremy- I realize that you can only see 8 bits, but you should be able to visualize 8 bits worth of data from different ranges of the 65000 gray levels in the image. I can easily look at 8 bits of bright pixels, but I can’t look at 8 bits of medium brightness or 8 bits of just over background information. The dim stuff is overwhelmed by the saturated bright pixels if you simply contrast the image to bring up those values, even though they are plenty bright relative to background.

Dr. David Knecht
Professor , Department of Molecular and Cell Biology
University of Connecticut
91 N. Eagleville Rd.
U-3125
Storrs, CT 06269-3125
860-486-2200

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