[Insight-users] Multi-threading for K-Means classification

[Kishore Mosaliganti]

Hi Anu,

You can consider downsampling your data (choose 5% of the points randomly)
and then initialize the k-means procedure on the full dataset with the new
means estimated from sampling.

I think this can speed up your processing to under 15s.

Also, I hope you are using the Kd-treee based k-means procedure.

Kishore



On Sun, Mar 7, 2010 at 2:59 PM, Anu <anuradha.tcs at gmail.com> wrote:

> yes, I have built the project in release mode and it takes around 1-2 mins
> for a 256*256*300 data.
> But i need this  to be more efficient.
>
> On Sun, Mar 7, 2010 at 5:51 AM, Luis Ibanez <luis.ibanez at kitware.com>wrote:
>
>> Hi Anu,
>>
>> I'm not aware of a multi-threaded version of this filter.
>>
>>
>> 1) What computation time are you observing ?
>>
>>
>> 2) Are you compiling your project in "Release" mode ?
>>    The compilation mode "Debug" versus "Release"
>>    can easily make a 10x difference in computation time
>>    for some ITK filters.
>>
>>
>>
>>   Regards,
>>
>>
>>          Luis
>>
>>
>> -------------------------------------------------------
>>  On Fri, Mar 5, 2010 at 4:57 AM, Anu <anuradha.tcs at gmail.com> wrote:
>> > Hi,
>> >
>> > Am trying to apply segmentation using k-Means clustering for CT images.
>> I
>> > have used itkScalarImageKmeansImageFilter for a 3D dataset (512 * 512 *
>> 400)
>> > of CT volume. It works fine but the processing speed is very slow on my
>> > intel dual core 2G RAM. Is there a multithreaded version already avaible
>> in
>> > the library or  any other faster algorithm option available?
>> > Thanks
>> > Anu.
>> >
>> >
>> >
>>  > _____________________________________
>> > 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
>> >
>> >
>>
>
>
>
> --
> Cheers !!
> Anu..
>
> _____________________________________
> 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
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20100310/ee9da790/attachment.htm>