Thanks Carlos, fully clarified now. I've checked and all values I find
are always <=0.
That note,
"This class returns the negative mutual information value."
should make it to the otb manual as well.

Agus


On Mon, Sep 7, 2015 at 11:15 AM, Carlos Hernando
<[email protected]> wrote:
> I think that it uses the Mattes version of MI, which provides negative
> values as output:
> http://www.itk.org/Doxygen/html/classitk_1_1MattesMutualInformationImageToImageMetric.html
>
>
> Best regards
> Carlos
>
> El viernes, 4 de septiembre de 2015, 17:38:17 (UTC+2), alobo escribió:
>>
>> No, they are not. Could the metric be centered around the mean?
>> Somebody with enough knowledge could check the code? Should I fill a
>> bug report?
>> Agus
>>
>> On Fri, Sep 4, 2015 at 4:57 PM, Jordi Inglada <[email protected]> wrote:
>> > Hi Agus,
>> >
>> > Maybe a silly question: are these negative values large?
>> >
>> > This could be due to numerical precision errors in the estimation, since
>> > the entropy itself is estimated using histograms.
>> >
>> > Jordi
>> >
>> > Agustin Lobo <[email protected]> wrote:
>> >>
>> >> otbcli_FineRegistration using MI as metric produces negative values
>> >> (saved to the 3rd layer of the result of otbcli_FineRegistration).
>> >>
>> >> For example, in
>> >> otbcli_FineRegistration -ref StereoFixed.png -sec StereoMoving.png
>> >> -out FineRegistrationMI.tif -erx 2 -ery 2 -mrx 3 -mry 3 -m MI
>> >>
>> >> The 3rd "band" of FineRegistrationMI.tif has negative values.
>> >>
>> >> As
>> >> H(A+B) <=H(A)+H(B)
>> >> and
>> >> MI = H(A)+H(B)-H(A+B)
>> >>
>> >> MI cannot be <0
>> >>
>> >> Any idea of what happens here?
>> >>
>> >> Thanks
>> >>
>> >> Agus
>> >>
>> >> --
>> >
>> > --
>> > --
>> > Check the OTB FAQ at
>> > http://www.orfeo-toolbox.org/FAQ.html
>> >
>> > You received this message because you are subscribed to the Google
>> > Groups "otb-users" group.
>> > To post to this group, send email to [email protected]
>> > To unsubscribe from this group, send email to
>> > [email protected]
>> > For more options, visit this group at
>> > http://groups.google.com/group/otb-users?hl=en
>> > ---
>> > You received this message because you are subscribed to the Google
>> > Groups "otb-users" group.
>> > To unsubscribe from this group and stop receiving emails from it, send
>> > an email to [email protected].
>> > For more options, visit https://groups.google.com/d/optout.
>
> --
> --
> Check the OTB FAQ at
> http://www.orfeo-toolbox.org/FAQ.html
>
> You received this message because you are subscribed to the Google
> Groups "otb-users" group.
> To post to this group, send email to [email protected]
> To unsubscribe from this group, send email to
> [email protected]
> For more options, visit this group at
> http://groups.google.com/group/otb-users?hl=en
> ---
> You received this message because you are subscribed to the Google Groups
> "otb-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> For more options, visit https://groups.google.com/d/optout.

-- 
-- 
Check the OTB FAQ at
http://www.orfeo-toolbox.org/FAQ.html

You received this message because you are subscribed to the Google
Groups "otb-users" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/otb-users?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"otb-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to