[Insight-users] Texture feature extraction

[Kim, Eun Young]

Hello Dženan,

I have investigated a little bit about the texture, including GLCM, and realized it is not a trivial problem. The texture feature is indeed attractive, however as you already realized that it is not realistic to use in the segmentation process as it is in ITK. (Our team also had been tried to improve the ITK filter.) There have been a lots of trials making it faster besides ITK, but mostly from the 'hardware'. This paper is one of the approach to enhance the computation time (and might help you to understand):

Texture, I., & Extraction, F. (2009). MSc THESIS Implementing Texture Feature Extraction Algorithms on FPGA. Electrical Engineering.

I have looked at several other articles a while ago, and remember that most of the solution/research were based on hardware.
As far as I know, there is no really good improvement in algorithmic way yet.
Please let me know if I have a wrong perspective about textures. I would love to use textures in my work as well. :)

Thank you.

Regina (Eun young)


On Feb 10, 2012, at 8:18 AM, Dženan Zukić wrote:

After finding out how GLCM is used to extract features, I have created a new wiki example<http://www.itk.org/Wiki/ITK/Examples/Statistics/TextureFeatures>. 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<http://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/20120210/50af0861/attachment.htm>