Is it possible to reduce contrast in certain region of an image

I am doing an iris recognition project, and I am trying to come up with a way to use a same threshold for edge detection to get the iris boundary. The problem is, with the same threshold, there is an extra edge within the iris region that affect the result of the hough transform. Is there anyway to make it so the contrast in the iris region is uniform so there won't be any extra edge?

Answer 1

[KALYAN ACHARJYA]

Decide the region of Interest Like Draw an ellipse over the image to specify the region of interest, using the imellipse function. More detail here

Then Apply imcontrast adjustment function.

ROI=imellipse(gca,[Four Co-ordinates value]); 
%Check the approximate values of coordinates using imtool

Answer 2

[Image Analyst]

Not exactly sure what you want to do from your diagrams. But there is a huge amount of literature on analyzing irises and I'm sure people have already solved the problem of identifying and separating the sclera, iris, and pupil. You just simply have to follow one of those published, perfected solutions.

http://www.visionbib.com/bibliography/contentspeople.html#Face%20Recognition,%20Detection,%20Tracking,%20Gesture%20Recognition,%20Fingerprints,%20Biometrics

[Lanceric Tse]

bloody helpful aren't you mate

[Image Analyst]

Well, I'm #2 on the list.

You'll find virtually every paper on image processing ever published on this list. It's a very good resource. Since I haven't done exactly whatever you want, I don't have code for it and whatever hack I, or others here, come up with would probably not be as good as those of people who have studied this for months and written a peer reviewed papers on it. Not sure if you've chosen to intentionally avoid published papers in favor of Answers, but that may not be your best route to a solution. But keep monitoring this post and you may find other solutions that are better than the published ones.

[Image Analyst]

Well, I'm #2 on the list.

You'll find virtually every paper on image processing ever published on this list. It's a very good resource. Since I haven't done exactly whatever you want, I don't have code for it and whatever hack I, or others here, come up with would probably not be as good as those of people who have studied this for months and written a peer reviewed papers on it. Not sure if you've chosen to intentionally avoid published papers in favor of Answers, but that may not be your best route to a solution. But keep monitoring this post and you may find other solutions that are better than the published ones.