7
$\begingroup$

palm

I am trying to figure out a way to find out the vein pattern in the hand.. Please could someone suggest ways to do the same.? I know first i would need to find the ROI in the image and perform further processing .. Can gabor filter be used as one of the steps ?

Thanks

$\endgroup$
  • $\begingroup$ See related question on SO: stackoverflow.com/questions/8694534/iphone-camera-application $\endgroup$ – Paul R Jan 19 '12 at 15:35
  • $\begingroup$ my question is what is the programming languge for those code becuse i try to apply that on MATLAB but it dos'nt work and if there any equivelant code in MATLAB. thanks zainab $\endgroup$ – user4798 Jun 12 '13 at 9:59
6
$\begingroup$

Ridge filtering seems like a possible processing. I have not tried to clean up the non-veins ridges; perhaps there would not be need for that step in more constrained acquisition conditions.

In Mathematica:

img = ColorConvert[Import["http://i.stack.imgur.com/5og5Z.jpg"], "Grayscale"];
r = RidgeFilter[ColorNegate@img, 3];
Binarize[ImageAdjust@r, .03] // Thinning

enter image description here

$\endgroup$
  • $\begingroup$ is there an implementation of ridge filtering in MATLAB? $\endgroup$ – vini Jan 18 '12 at 16:24
  • $\begingroup$ There does not seem to be any implementation shipping with matlab. However, Google reveals user contributions to the matlab central exchange. According to the Mathematica documentation (reference.wolfram.com/mathematica/ref/RidgeFilter.html), ones needs to compute "the main principal curvature orthogonal to a ridge as given by the main negative eigenvalue of the Hessian matrix." $\endgroup$ – Matthias Odisio Jan 18 '12 at 16:54
  • $\begingroup$ I find this concept a little hard to grasp ! $\endgroup$ – vini Jan 18 '12 at 17:30
  • $\begingroup$ One reference would be "Front-End Vision and Multi-Scale Image Analysis", by Bart M. Haar Romeny $\endgroup$ – Matthias Odisio Jan 18 '12 at 17:47
  • $\begingroup$ will search and let u knw :) $\endgroup$ – vini Jan 18 '12 at 18:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.