[Insight-users] SetNumberOfSpatialSamples has no effect on computation time???

[Luis Ibanez]

Hi  Pinpress,

Thanks for your clarification.

---

A change from 10% of samples to 100% of samples should be
noticeable in the time that it takes to compute the registration.


A)  How big are your images ?

B)  How many pixels do they have ?

C)  How many iterations did your optimizer ran ?

D)  What Transform are you using ?



something else might be at play here....



     Luis


------------------------------------------------------------------
On Wed, May 26, 2010 at 2:38 PM, pinpress <sb_ji at yahoo.com> wrote:
>
> Hi Luis,
>
> Thanks for the reply. Yes, I was sure that I compiled in Release mode, and I
> have set the two Cmake Flags so that to use optimized registration
> techniques. The "SetNumberOfSpatialSamples ()" function does have a very
> noticeable effect on the performance on Mattes Registration technique. I
> have timed the registration (using "time" in Linux) using normalized mutual
> information or correlation coefficient image similarity metrics to perform
> the registration. By changing the # of samples to only 10% or 100% of the
> input # of voxels, the computation times were the same.
>
> Therefore, I suspect this may be because the SetNumberOfSpatialSamples()
> function is by-passed somewhere down the path of implementation?
>
> Thanks for looking into this!
>
>
> Luis Ibanez wrote:
>>
>> Hi Pinpress,
>>
>> You will have to increase by a lot the number of samples
>> in order to notice an effect in the computation time.
>>
>> If your real goal is to speed up the registration process,
>> the most important thing to do is to make sure that
>> you are compiling for "Release" mode, and to check
>> if all the cores of your processor are being used at
>> run time.
>>
>>
>>      Regards,
>>
>>
>>             Luis
>>
>>
>>
>> --------------------------------
>> On Fri, May 7, 2010 at 10:43 AM, pinpress <sb_ji at yahoo.com> wrote:
>>>
>>> Dear all,
>>>
>>> I am performing 3D image registration with data set size of
>>> 256-by-256-by-124. Using normalized MI, or correlation coefficient, it's
>>> really slow. Therefore, I have purposefully set
>>> SetNumberOfSpatialSamples( )
>>> to a small number, e.g., 10% of the total number of voxels. To my dismay,
>>> it
>>> does not seem to have any effect on the computation time.
>>>
>>> Can I ask what's the reason? I have already set the optimization flags to
>>> be
>>> "ON" when compiling ITK using cmake. Therefore, this function does
>>> influence
>>> computation time when I use Mattes MI or Mean Squares Difference as the
>>> similarity measure.
>>>
>>> Please help. Thanks!
>>> --
>>> View this message in context:
>>> http://old.nabble.com/SetNumberOfSpatialSamples-has-no-effect-on-computation-time----tp28486979p28486979.html
>>> Sent from the ITK - Users mailing list archive at Nabble.com.
>>>
>>> _____________________________________
>>> 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
>>
>>
>
> --
> View this message in context: http://old.nabble.com/SetNumberOfSpatialSamples-has-no-effect-on-computation-time----tp28486979p28684658.html
> Sent from the ITK - Users mailing list archive at Nabble.com.
>
> _____________________________________
> 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
>