Threshold ... why does it not work exactly ?

[bal]

Hi All

Sorry a little new to this ......

When I use ImageJ: Adjust: Threshold on 16-bit medical image volume dataset and set the threshold to a specific upper and lower limit value (when set to default and red or b/w in the tabs) it appears the actual thresholding is not based exactly on the exact limits I have set but appear to be based on a wider range than the ones I have set.

Should the simple threshold be based just on my upper and lower limits as I input ?

Please let me know if I am doing this wrong.

kind regards
bal

[Michael Schmid]

Hi Bal,

your observation is correct, the colors do not exactly reflect the  
threshold settings with full accuracy of 16-bit or float images.

The reason:
The displayed image has only 8 bits of resolution, and the red  
display is simply done by updating the lookup table, which is very  
fast (this was especially important in the early times of ImageJ),  
but has only 8 bits of precision.

An accurate solution would be clearly desired these days of fast  
computers, but it would require quite some effort to make it work.

So far my view of the problem, others might disagree...

Michael
________________________________________________________________

On 29 Nov 2010, at 18:08, Bal Sanghera wrote:

> Hi All
>
> Sorry a little new to this ......
>
> When I use ImageJ: Adjust: Threshold on 16-bit medical image volume  
> dataset and set the threshold to a specific upper and lower limit  
> value (when set to default and red or b/w in the tabs) it appears  
> the actual thresholding is not based exactly on the exact limits I  
> have set but appear to be based on a wider range than the ones I  
> have set.
>
> Should the simple threshold be based just on my upper and lower  
> limits as I input ?
>
> Please let me know if I am doing this wrong.
>
> kind regards
> bal

[bal]

thanks for your reply

regards
bal

_____________________________

Bal Sanghera PhD
Paul Strickland Scanner Centre
Mount Vernon Hospital
Northwood. HA6 2RN.
Tel : [phone]
Fax:  [phone]
_____________________________


________________________________________
From: ImageJ Interest Group [[hidden email]] On Behalf Of Michael Schmid [[hidden email]]
Sent: 29 November 2010 06:00 PM
To: [hidden email]
Subject: Re: Threshold ... why does it not work exactly ?

Hi Bal,

your observation is correct, the colors do not exactly reflect the
threshold settings with full accuracy of 16-bit or float images.

The reason:
The displayed image has only 8 bits of resolution, and the red
display is simply done by updating the lookup table, which is very
fast (this was especially important in the early times of ImageJ),
but has only 8 bits of precision.

An accurate solution would be clearly desired these days of fast
computers, but it would require quite some effort to make it work.

So far my view of the problem, others might disagree...

Michael
________________________________________________________________

On 29 Nov 2010, at 18:08, Bal Sanghera wrote:

> Hi All
>
> Sorry a little new to this ......
>
> When I use ImageJ: Adjust: Threshold on 16-bit medical image volume
> dataset and set the threshold to a specific upper and lower limit
> value (when set to default and red or b/w in the tabs) it appears
> the actual thresholding is not based exactly on the exact limits I
> have set but appear to be based on a wider range than the ones I
> have set.
>
> Should the simple threshold be based just on my upper and lower
> limits as I input ?
>
> Please let me know if I am doing this wrong.
>
> kind regards
> bal

********************************************************************************************************************

This message may contain confidential information. If you are not the intended recipient please inform the
sender that you have received the message in error before deleting it.
Please do not disclose, copy or distribute information in this e-mail or take any action in reliance on its contents:
to do so is strictly prohibited and may be unlawful.

Thank you for your co-operation.

NHSmail is the secure email and directory service available for all NHS staff in England and Scotland
NHSmail is approved for exchanging patient data and other sensitive information with NHSmail and GSi recipients
NHSmail provides an email address for your career in the NHS and can be accessed anywhere
For more information and to find out how you can switch, visit www.connectingforhealth.nhs.uk/nhsmail

********************************************************************************************************************

[Daniel Kalthoff]

A workaround that might be helpful is via Process > Math > Macro.
An expression like "v=v*(v>10)" would set all pixels to zero if they are smaller or equal to 10, since in this case the bracket evaluates to false (equivalent to 0). You can play around with combinations of upper / lower thresholds via logical operators to achieve what you (might) want.

Best regards,

Daniel

[Daniel Kalthoff]

A workaround that might be helpful is via Process > Math > Macro.
An expression like "v=v*(v>10)" would set all pixels to zero if they are smaller or equal to 10, since in this case the bracket evaluates to false (equivalent to 0). You can play around with combinations of upper / lower thresholds via logical operators to achieve what you (might) want.

Best regards,

Daniel

[Bingley]

Sorry if this is a silly question. I have spent quite a lot of time but
have not be able to find a more detailed, step by step guide on getting
a complex (developed as a seperate project in Eclipse with 50+ classes)
plugin to work on ImageJ Applet. Please has anyone come across any
tutorial of such kind and I would be very grateful if you could let me
know.

I did find the following however it is not detailed enough for me and
just thought to put it here for information as it might be useful for
someone else.

http://bij.isi.uu.nl/applet.htm

Many thanks,
Bingley