Help for the Haralic texture feature extraction

35 views
Skip to first unread message

get...@gmail.com

unread,
Jan 21, 2015, 6:14:15 AM1/21/15
to jfeat...@googlegroups.com
So much thanks Dr. Fraz for your quick response!
currently I am working  Image processing . I decided to use Haralic texture feature extraction
on jfeaturelib 1.6.1, but every class doesn't have "main  method"! so what i shall do?
I don't know where to start!
i have tried the "Moments.java" class to modify
but i couldn't !
any help please!
thanks in advance!

Franz

unread,
Jan 21, 2015, 9:09:28 AM1/21/15
to jfeat...@googlegroups.com
Hi,

https://code.google.com/p/jfeaturelib/wiki/HowTo is a good starting point for how to extract features from images.
The main method you're looking for can be found in de.lmu.ifi.dbs.jfeaturelib.utils.Extractor

Regards
Franz

get...@gmail.com

unread,
Jan 22, 2015, 7:37:28 AM1/22/15
to jfeat...@googlegroups.com


Dear  Dr. Franz Graf,
 First and foremost I would like  to thanks!
But,it is a little more complex!
i just added  jfeaturelib 1.6.1.jar with depende. to my Netbeans IDE
I have segmented my image using k-means,Region growing and Fuczy C means segmentation Mechanism
now what i need is that, when i click on one region of the segmented image it "Calculates the basic statistical moments of an image like mean, std_dev,  skewness and kurtosis" etc of the given point. I have gone through the "moments.java" and "Haralic.java"
but i am getting some problems would you help me please?
 
Haralick.java
Moments.java

Franz Graf

unread,
Jan 23, 2015, 1:01:58 AM1/23/15
to jfeat...@googlegroups.com
Hi,

Ah I see. So what you want is probably the extraction from masked areas of the image.
JFeatureLib provides masking only for a subset of features and unfortunately Haralick does not yet support masking.

The only thing you can do is the extraction from a bounding box around (or enclosed by) the region of interest. If this is acceptable, you can extract an image processor (from ImageJ) and call the Haralick class directly. In this case you have to deal a little bit with imageJ.

Hope this helps.
Good luck
Franz
Dr. Franz Graf
http://www.Locked.de
Reply all
Reply to author
Forward
0 new messages