Discussion:
SIFT/feature based registration in ITK?
David Doria
2009-11-16 22:16:18 UTC
Permalink
I see a whole bunch of intensity based registration algorithms in the
software guide, but I don't see any feature based registration. Am I missing
them? The required parts would be feature extraction, feature
correspondence, and then feature based transformation estimation.

I saw this paper:
http://www.insight-journal.org/browse/publication/207

which seems to talk about using the existing 2D SIFT, etc, but can someone
point me to where this lives/an example?

Thanks!

David
David Doria
2009-11-19 20:40:17 UTC
Permalink
Someone mentioned off list that they are slowly working on writing
SIFT for ITK - has there really been no past effort like this? Is the
idea that ITK deals mainly with medical images and intensity based
registration is typically fine for those?

Thanks,

David
_____________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Kitware offers ITK Training Courses, for more information visit:
http://www.kitware.com/products/protraining.html

Please keep messages on-topic and check the ITK FAQ at:
http://www.itk.org/Wiki/ITK_FAQ

Follow this link to subscribe/unsubscribe:
http://www.itk.org/mailman/listinfo/insight-users
Torsten Rohlfing
2009-11-19 20:45:22 UTC
Permalink
Another reason for lack of motivation to implement SIFT in ITK might
also be that SIFT is patented.
Post by David Doria
Someone mentioned off list that they are slowly working on writing
SIFT for ITK - has there really been no past effort like this? Is the
idea that ITK deals mainly with medical images and intensity based
registration is typically fine for those?
Thanks,
David
--
Torsten Rohlfing, PhD SRI International, Neuroscience Program
Senior Research Scientist 333 Ravenswood Ave, Menlo Park, CA 94025
Phone: ++1 (650) 859-3379 Fax: ++1 (650) 859-2743
torsten-RAQ/***@public.gmane.org http://www.stanford.edu/~rohlfing/

"Though this be madness, yet there is a method in't"
Rupin
2009-11-19 23:43:40 UTC
Permalink
Are you not allowed to use patented methods for non-commercial research !?!

Rupin

On Thu, 19 Nov 2009 12:45:22 -0800, Torsten Rohlfing
Post by Torsten Rohlfing
Another reason for lack of motivation to implement SIFT in ITK might
also be that SIFT is patented.
Post by David Doria
Someone mentioned off list that they are slowly working on writing
SIFT for ITK - has there really been no past effort like this? Is the
idea that ITK deals mainly with medical images and intensity based
registration is typically fine for those?
Thanks,
David
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
_____________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Kitware offers ITK Training Courses, for more information visit:
http://www.kitware.com/products/protraining.html

Please keep messages on-topic and check the ITK FAQ at:
http://www.itk.org/Wiki/ITK_FAQ

Follow this link to subscribe/unsubscribe:
http://www.itk.org/mailman/listinfo/insight-users
Bill Lorensen
2009-11-20 04:24:54 UTC
Permalink
There is recent patent law that muddies the term non-commercial use.
For example, if you are at a University your use of a patent may well
be considered commercial use.

See http://www.freedom-to-tinker.com/blog/felten/experimental-use-exception-evaporating
for example.
Post by Rupin
Are you not allowed to use patented methods for non-commercial research !?!
Rupin
On Thu, 19 Nov 2009 12:45:22 -0800, Torsten Rohlfing
Post by Torsten Rohlfing
Another reason for lack of motivation to implement SIFT in ITK might
also be that SIFT is patented.
Post by David Doria
Someone mentioned off list that they are slowly working on writing
SIFT for ITK - has there really been no past effort like this? Is the
idea that ITK deals mainly with medical images and intensity based
registration is typically fine for those?
Thanks,
David
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
_____________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html
http://www.kitware.com/products/protraining.html
http://www.itk.org/Wiki/ITK_FAQ
http://www.itk.org/mailman/listinfo/insight-users
_____________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Kitware offers ITK Training Courses, for more information visit:
http://www.kitware.com/products/protraining.html

Please keep messages on-topic and check the ITK FAQ at:
http://www.itk.org/Wiki/ITK_FAQ

Follow this link to subscribe/unsubscribe:
http://www.itk.org/mailman/listinfo/insight-users
Michael Mai
2009-11-20 07:33:27 UTC
Permalink
See http://www.cmake.org/Wiki/ITK_Patent_Bazaar for answer.
Summarized: Unfortunately no, even non-commercial research have to obey patents.

Michael

-----Ursprüngliche Nachricht-----
Von: insight-users-bounces-***@public.gmane.org [mailto:insight-users-bounces-***@public.gmane.org] Im Auftrag von Rupin
Gesendet: Freitag, 20. November 2009 00:44
An: insight-users-***@public.gmane.org
Betreff: Re: [Insight-users] SIFT/feature based registration in ITK?

Are you not allowed to use patented methods for non-commercial research !?!

Rupin

On Thu, 19 Nov 2009 12:45:22 -0800, Torsten Rohlfing
Post by Torsten Rohlfing
Another reason for lack of motivation to implement SIFT in ITK might
also be that SIFT is patented.
Post by David Doria
Someone mentioned off list that they are slowly working on writing
SIFT for ITK - has there really been no past effort like this? Is the
idea that ITK deals mainly with medical images and intensity based
registration is typically fine for those?
Thanks,
David
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
_____________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Kitware offers ITK Training Courses, for more information visit:
http://www.kitware.com/products/protraining.html

Please keep messages on-topic and check the ITK FAQ at:
http://www.itk.org/Wiki/ITK_FAQ

Follow this link to subscribe/unsubscribe:
http://www.itk.org/mailman/listinfo/insight-users
_____________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Kitware offers ITK Training Courses, for more information visit:
http://www.kitware.com/products/protraining.html

Please keep messages on-topic and check the ITK FAQ at:
http://www.itk.org/Wiki/ITK_FAQ

Follow this link to subscribe/unsubscribe:
http://www.itk.org/mailman/listinfo/insight-users
Luis Ibanez
2009-11-23 01:09:18 UTC
Permalink
As Bill and Michael have pointed out,

The "Experimental" exception for the exclusive rights of Patents
(in the US) applies only:

"for amusement, to satisfy idle curiosity,
or strictly for philosophical inquiry."

See:

http://www.bakerbotts.com/infocenter/publications/detail.aspx?id=b7930f1d-b945-4f95-b825-fa9ac70c16af

and

http://books.google.com/books?id=-gLuY2rBU9oC&pg=RA2-PA9-IA466&lpg=RA2-PA9-IA466&dq=for+amusement,+to+satisfy+idle+curiosity&source=bl&ots=qQdZEW6lDu&sig=VyoBqK6AHJ4iEUKsRbyeXPJLjBg&hl=en&ei=idoJS4zHDpCZlAeYhaGhDA&sa=X&oi=book_result&ct=result&resnum=4&ved=0CBAQ6AEwAw#v=onepage&q=for%20amusement%2C%20to%20satisfy%20idle%20curiosity&f=false


A part from that,
you are not allowed to:

* Use
* Make
* Sell
* Offer for sale
* Import

any embodiment of an invention that is protected by a Patent,
without the permission of the Patent holder.


More on the Research Exemption for Patents at

http://en.wikipedia.org/wiki/Research_exemption


Note also that Patents have only national jurisdiction.
That is, the patent in question here:

"A method and apparatus for identifying scale invariant features in
an image and a further method and apparatus for using such scale
invariant features to locate an object in an image"

http://www.google.com/patents?vid=USPAT6711293

only apply to the United States.



Therefore, if you are outside of the US, you can make, use,
sell, and offer for sale any implementation of this US patented
method.

Beware however,
that is common for companies to patent the same invention
in multiple countries. Therefore, you still have to check with
the patent database of your respective country.


For more on Patents and how they obstruct the progress of science
and technology you may want to read.

"Patent Failure"
"How Judges,Bureaucrats, and Lawyers Put Innovators at Risk"
by James Bessen and Michael J. Meurer
Princeton University Press
http://www.researchoninnovation.org/dopatentswork/

and

"Math you can't use:"
"Patents, Copyright, and Software "
http://www.amazon.com/Math-You-Cant-Use-Copyright/dp/0815749422

and

"The Public Domain"
"Enclosing the Commons of the Mind"
by James Boyle
http://www.thepublicdomain.org/



BTW,
The US Supreme Court is currently reviewing the criteria that the
US Patent Office should use for granting patents.
http://online.wsj.com/article/SB10001424052748704431804574537481229336114.html

<quote>

The most telling moment in the Bilski argument was when Justice Breyer
asked how the balance should be struck between granting patents for
methods that applied to machines as opposed to methods that apply to
how information is used. "I don't know," he answered. "And I don't
know whether across the board or in this area or that area patent
protection would do no harm or more harm than good."

Likewise, Justice Sotomayor said she couldn't predict the result if
the court tried to clarify what can be patented and what can't. "I
have no idea what the limits of that ruling will impose in the
computer world, in the biomedical world."

Such humility is rare at the Supreme Court, but as the justices come
to a decision in this case, they should remember above all that legal
uncertainty about intellectual property has real costs. For now, the
most innovative parts of our economy bear the burden of uncertainty,
with no one knowing for sure who owns what rights to which ideas,
inventions or discoveries.

</quote>


It is not surprising that Patent Systems were abolished in several
countries in the past: for example in The Netherlands in 1869.

By the time of the American Revolution; England was the only country
where a Patent system was in place. It was common knowledge that
the creation of Monopolies was detrimental for the Economy, particularly
for a market economy.



In August 1813, Thomas Jefferson wrote:

"I am informed, that England was, until we copied her, the only
country on earth which ever, by a general law, gave a legal right to
the exclusive use of an idea. In some other countries it is sometimes
done, in a great case, and by a special and personal act, but,
generally speaking, other nations have thought that these monopolies
produce more embarrassment than advantage to society; and it may be
observed that the nations which refuse monopolies of invention, are as
fruitful as England in new and useful devices."

http://press-pubs.uchicago.edu/founders/documents/a1_8_8s12.html



Regards,



Luis


----------------------------------------------------------------------------------------------------------------------
Post by Michael Mai
See http://www.cmake.org/Wiki/ITK_Patent_Bazaar for answer.
Summarized: Unfortunately no, even non-commercial research have to obey patents.
Michael
-----Ursprüngliche Nachricht-----
Gesendet: Freitag, 20. November 2009 00:44
Betreff: Re: [Insight-users] SIFT/feature based registration in ITK?
Are you not allowed to use patented methods for non-commercial research !?!
Rupin
On Thu, 19 Nov 2009 12:45:22 -0800, Torsten Rohlfing
Post by Torsten Rohlfing
Another reason for lack of motivation to implement SIFT in ITK might
also be that SIFT is patented.
Post by David Doria
Someone mentioned off list that they are slowly working on writing
SIFT for ITK - has there really been no past effort like this? Is the
idea that ITK deals mainly with medical images and intensity based
registration is typically fine for those?
Thanks,
David
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
_____________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html
http://www.kitware.com/products/protraining.html
http://www.itk.org/Wiki/ITK_FAQ
http://www.itk.org/mailman/listinfo/insight-users
_____________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html
http://www.kitware.com/products/protraining.html
http://www.itk.org/Wiki/ITK_FAQ
http://www.itk.org/mailman/listinfo/insight-users
_____________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Kitware offers ITK Training Courses, for more information visit:
http://www.kitware.com/products/protraining.html

Please keep messages on-topic and check the ITK FAQ at:
http://www.itk.org/Wiki/ITK_FAQ

Follow this link to subscribe/unsubscribe:
http://www.itk.org/mailman/listinfo/insight-users
amit kumar
2009-11-23 06:26:17 UTC
Permalink
Several of us feel strongly about how software patents obstruct research.
Interested to hear your thoughts on what the community could do to stymie
this practice or circumvent fuzzy generic patents ?

1. Can we maintain a list of list of candidate patents where alternate
methods are sought ?.
See http://en.wikipedia.org/wiki/Marching_cubes
Marching tetrahedrons came out of such a response from the computer graphics
community.

2. Several applicants seek to patent specific applications of published
techiniques.
Are there patents obvious ?. What can we do about them ?

Hope we can identify some actionable baby steps..

regards,
ak
Post by Luis Ibanez
As Bill and Michael have pointed out,
The "Experimental" exception for the exclusive rights of Patents
"for amusement, to satisfy idle curiosity,
or strictly for philosophical inquiry."
http://www.bakerbotts.com/infocenter/publications/detail.aspx?id=b7930f1d-b945-4f95-b825-fa9ac70c16af
and
http://books.google.com/books?id=-gLuY2rBU9oC&pg=RA2-PA9-IA466&lpg=RA2-PA9-IA466&dq=for+amusement,+to+satisfy+idle+curiosity&source=bl&ots=qQdZEW6lDu&sig=VyoBqK6AHJ4iEUKsRbyeXPJLjBg&hl=en&ei=idoJS4zHDpCZlAeYhaGhDA&sa=X&oi=book_result&ct=result&resnum=4&ved=0CBAQ6AEwAw#v=onepage&q=for%20amusement%2C%20to%20satisfy%20idle%20curiosity&f=false
A part from that,
* Use
* Make
* Sell
* Offer for sale
* Import
any embodiment of an invention that is protected by a Patent,
without the permission of the Patent holder.
More on the Research Exemption for Patents at
http://en.wikipedia.org/wiki/Research_exemption
Note also that Patents have only national jurisdiction.
"A method and apparatus for identifying scale invariant features in
an image and a further method and apparatus for using such scale
invariant features to locate an object in an image"
http://www.google.com/patents?vid=USPAT6711293
only apply to the United States.
Therefore, if you are outside of the US, you can make, use,
sell, and offer for sale any implementation of this US patented
method.
Beware however,
that is common for companies to patent the same invention
in multiple countries. Therefore, you still have to check with
the patent database of your respective country.
For more on Patents and how they obstruct the progress of science
and technology you may want to read.
"Patent Failure"
"How Judges,Bureaucrats, and Lawyers Put Innovators at Risk"
by James Bessen and Michael J. Meurer
Princeton University Press
http://www.researchoninnovation.org/dopatentswork/
and
"Math you can't use:"
"Patents, Copyright, and Software "
http://www.amazon.com/Math-You-Cant-Use-Copyright/dp/0815749422
and
"The Public Domain"
"Enclosing the Commons of the Mind"
by James Boyle
http://www.thepublicdomain.org/
BTW,
The US Supreme Court is currently reviewing the criteria that the
US Patent Office should use for granting patents.
http://online.wsj.com/article/SB10001424052748704431804574537481229336114.html
<quote>
The most telling moment in the Bilski argument was when Justice Breyer
asked how the balance should be struck between granting patents for
methods that applied to machines as opposed to methods that apply to
how information is used. "I don't know," he answered. "And I don't
know whether across the board or in this area or that area patent
protection would do no harm or more harm than good."
Likewise, Justice Sotomayor said she couldn't predict the result if
the court tried to clarify what can be patented and what can't. "I
have no idea what the limits of that ruling will impose in the
computer world, in the biomedical world."
Such humility is rare at the Supreme Court, but as the justices come
to a decision in this case, they should remember above all that legal
uncertainty about intellectual property has real costs. For now, the
most innovative parts of our economy bear the burden of uncertainty,
with no one knowing for sure who owns what rights to which ideas,
inventions or discoveries.
</quote>
It is not surprising that Patent Systems were abolished in several
countries in the past: for example in The Netherlands in 1869.
By the time of the American Revolution; England was the only country
where a Patent system was in place. It was common knowledge that
the creation of Monopolies was detrimental for the Economy, particularly
for a market economy.
"I am informed, that England was, until we copied her, the only
country on earth which ever, by a general law, gave a legal right to
the exclusive use of an idea. In some other countries it is sometimes
done, in a great case, and by a special and personal act, but,
generally speaking, other nations have thought that these monopolies
produce more embarrassment than advantage to society; and it may be
observed that the nations which refuse monopolies of invention, are as
fruitful as England in new and useful devices."
http://press-pubs.uchicago.edu/founders/documents/a1_8_8s12.html
Regards,
Luis
----------------------------------------------------------------------------------------------------------------------
Post by Michael Mai
See http://www.cmake.org/Wiki/ITK_Patent_Bazaar for answer.
Summarized: Unfortunately no, even non-commercial research have to obey
patents.
Post by Michael Mai
Michael
-----Ursprüngliche Nachricht-----
Im Auftrag von Rupin
Post by Michael Mai
Gesendet: Freitag, 20. November 2009 00:44
Betreff: Re: [Insight-users] SIFT/feature based registration in ITK?
Are you not allowed to use patented methods for non-commercial research
!?!
Post by Michael Mai
Rupin
On Thu, 19 Nov 2009 12:45:22 -0800, Torsten Rohlfing
Post by Torsten Rohlfing
Another reason for lack of motivation to implement SIFT in ITK might
also be that SIFT is patented.
Post by David Doria
Someone mentioned off list that they are slowly working on writing
SIFT for ITK - has there really been no past effort like this? Is the
idea that ITK deals mainly with medical images and intensity based
registration is typically fine for those?
Thanks,
David
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
_____________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html
http://www.kitware.com/products/protraining.html
http://www.itk.org/Wiki/ITK_FAQ
http://www.itk.org/mailman/listinfo/insight-users
_____________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html
http://www.kitware.com/products/protraining.html
http://www.itk.org/Wiki/ITK_FAQ
http://www.itk.org/mailman/listinfo/insight-users
_____________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html
http://www.kitware.com/products/protraining.html
http://www.itk.org/Wiki/ITK_FAQ
http://www.itk.org/mailman/listinfo/insight-users
Luis Ibanez
2009-11-23 00:09:19 UTC
Permalink
Hi David,

Your observation is correct.

ITK has a lot of support for intensity-based registration method,
but not much for feature-based methods.

This will be a prolific area for adding methods to ITK in the future.



Luis


------------------------------------------------------------------------
Post by David Doria
I see a whole bunch of intensity based registration algorithms in the
software guide, but I don't see any feature based registration. Am I missing
them? The required parts would be feature extraction, feature
correspondence, and then feature based transformation estimation.
http://www.insight-journal.org/browse/publication/207
which seems to talk about using the existing 2D SIFT, etc, but can someone
point me to where this lives/an example?
Thanks!
David
_____________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html
http://www.kitware.com/products/protraining.html
http://www.itk.org/Wiki/ITK_FAQ
http://www.itk.org/mailman/listinfo/insight-users
_____________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Kitware offers ITK Training Courses, for more information visit:
http://www.kitware.com/products/protraining.html

Please keep messages on-topic and check the ITK FAQ at:
http://www.itk.org/Wiki/ITK_FAQ

Follow this link to subscribe/unsubscribe:
http://www.itk.org/mailman/listinfo/insight-users
Loading...