Gabor filter can be simply defined as a Gaussian modulated sinusoid with a capability of multiresolution analysis. The extracted features represent the energy of each filter magnitude response.

More information can be found in section 6.8 of the PhD thesis on the use of Gabor filter for texture analysis in medical images: http://omar.alkadi.net/wp-content/uploads/Al-Kadi_PhD_Thesis.pdf

HI Ramiro,

plz go through the following link, it may help you

http://www.cse.iitm.ac.in/~sdas/vplab/courses/CV_DIP/PDF/Lect-Gabor_filt.pdf

and i hv listed some good research article on your interest plz find the attachments and go thorugh 

http://www.cs.technion.ac.il/users/wwwb/cgi-bin/tr-get.cgi/2005/CIS/CIS-2005-05.pdf

http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=1640626&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D1640626

Apply one level Gabor filter to the image,

Take all three filtered channel data (except the approximation channel), divide each channel matrix into for example 9 *9  blocks, calculate the mean and variance of each block, construct a feature vector, depending on your need, which contains the mean and variance information, for comparison with similar data in other images.

Hi, thank your answer. Inmediatly go to get the papers.