CORF uses as input the responses of center-on and center-off difference of Gaussians (DoG) operators. It achieves orientation selectivity by combining responses of DoG operators that are aligned in a colinear manner at a given orientation.
Besides orientation selectivity, the CORF operator achieves contrast invariant orientation tuning, cross orientation suppression and response saturation, typical properties of simple cells.
You are kindly invited to use this Matlab script and cite the following article:
George Azzopardi and Nicolai Petkov, "A CORF computational model of a simple cell that relies on LGN input outperforms the Gabor function model", Biological Cybernetics, vol.106(3), pp.177-189, 2012.
Paper [pdf]: http://link.springer.com/content/pdf/10.1007%2Fs00422-012-0486-6.pdf
Hi Hemo, could you kindly provide the statement that you are running and send me (email: geazzo<at>gmail<dot>com) the image file that you are trying to process?
can you please tell me why it doesn't work with me when I run it, it gives me an error (undefined function)]
Please send me paper
"Contour Detection by CORF Operator"
George Azzopardi, Nicolai Petkov
Artificial Neural Networks and Machine Learning – ICANN 2012
In 2012a to 2013b update code CORF.m at line 103 with this
mem = find(ismember(data.params,[polarity rho],'rows','legacy'),1);
Thanks Richard. I will update the script.
Working with 2013a. Added the parameter 'legacy' to ismember at line 103 to prevent an error.
Fixed a bug. The variable mem is now first initialised to .
Added a condition which checks if the data.params is empty
At line 103
Added a link to download the corresponding paper
I updated the Description by including the volume number and the page numbers of the citation to the paper on Biological Cybernetics