[Insight-users] Texture feature extraction

[Bradley Lowekamp]

Hello,

I am surprised about the large amount of interest in this type of filter.

Briefly comparing performance to the wiki example this is 6x faster single threaded, and it's multi-threaded so it can be like 100x with enough cores and it's streamable. 

I have placed what I had hacked together into github:

https://github.com/blowekamp/itkTextureAnalysis

I hopefully will get the testing and documentation into a respectable state this week.

Please use github for code comments and pull request for contributions.

Brief TODO:

Write documentation to describe what the filters do.
Write example like tests.
Write tests to verify computation, and parameters.
Improve the GLCM texture to use a moving histogram style algorithm.

Enjoy,
Brad


On Feb 10, 2012, at 10:14 AM, Johnson, Hans J wrote:

> Brad,
> 
> Eun Young (Regina) Kim is a graduate student of mine, and she would also be interested in this work.
> 
> Hans
> -- 
> Hans J. Johnson, Ph.D.
> hans-johnson at uiowa.edu
> Assistant Professor of Psychiatry
> University of Iowa Carver College of Medicine
> W278 GH, 200 Hawkins Drive
> Iowa City, Iowa 52242
> Phone:  319-353-8587
> 
> From: Bradley Lowekamp <blowekamp at mail.nih.gov>
> Date: Fri, 10 Feb 2012 09:34:36 -0500
> To: Dženan Zukić <dzenanz at gmail.com>
> Cc: Insight-users <insight-users at itk.org>
> Subject: Re: [Insight-users] (off-list) Texture feature extraction
> 
> Hello Dženan,
> 
> It appears that you are trying to compute texture attributes for each pixel. This is very much not what the texture filters was originally designed for. It was designed to extract texture attributes from a segmented region.
> 
> A couple months ago I prototyped a couple of filters to run this on a per-pixel basis, as your wiki example did. They worked to prove my point but I have not gotten the chance to return to this work. I am not guaranteeing the quality of the code...
> 
> I could extract them into a external ITK module and place them on github. Perhaps there you could fork it and continue work or other wise collaborate on this work?
> 
> Brad
> 
> 
> 
> On Feb 10, 2012, at 9:18 AM, Dženan Zukić wrote:
> 
>> After finding out how GLCM is used to extract features, I have created a new wiki example. However, the execution speed has really disappointed me.
>> 
>> The example is single-threaded, but even so 90 seconds for a small image (64x64x12) is too much. I wanted to use texture features along some other features for segmentation purposes, but really slow execution of this texture extraction technique has me reconsidering the usage of texture features.
>> 
>> If anyone has an idea how to significantly reduce execution time (in addition to the obvious parallelization), please let me know.
>> 
>> Regards,
>> Dženan
>> _____________________________________
>> 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
> 
> ========================================================
> Bradley Lowekamp  
> Medical Science and Computing for
> Office of High Performance Computing and Communications
> National Library of Medicine 
> blowekamp at mail.nih.gov
> 
> 
> 
> _____________________________________ 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 
> 
> Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521, is confidential and may be legally privileged.  If you are not the intended recipient, you are hereby notified that any retention, dissemination, distribution, or copying of this communication is strictly prohibited.  Please reply to the sender that you have received the message in error, then delete it.  Thank you.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20120214/4134cc74/attachment-0001.htm>