Sclera blood vessel enhancement on the UBIRIS database

Hi!

I'm trying to do sclera blood vessel enhancement on the UBIRIS color database having images of 600x800 pixels and I got stuck. I did sclera detection and achieved a pretty reasonable result (fig. 1), but when I try to do blood vessel enhancement all I get on the edge detected image are the boundary segments of the sclera and not the blood vessels inside it (fig. 2).

My first try was with a bank of Gabor filters according to this article: https://scholarworks.iupui.edu/bitstream/handle/1805/2096/Thesis_OneDoc_Thomas_FinalETD.pdf (page 61 in the PDF)

Then I tried with Frangi filters as some post on this forum says that they also had the problem with a bank of Gabor filters, but Frangi filters were working for them. Well, not for me.

Does anyone have an idea in what direction to go in order to have a well enhanced picture? For the article I posted I tried the following code:

I = rgb2gray(segmentedImage);
gaborArray = gabor([3,4,5],[0,90,180,270]); %maybe 0,90,180,270 or 0,45,90,135
[gaborMag, gaborPhase] = imgaborfilt(I,gaborArray);
f = zeros(size(I,1),size(I,2));
for i = 1:size(I,1)
    for j = 1:size(I,2)
        for k = 1:12
            f(i,j) = f(i,j) + gaborMag(i,j,k).^2;
        end
        f(i,j) = sqrt(f(i,j));
    end
end

maxh = max(f(:));

for i = 1:size(I,1)
    for j = 1:size(I,2)
        f(i,j) = f(i,j)/maxh;
    end
end

Am I not using correctly the orientations? Or should I combined the results of the 12 filters in a different way? The final normalization by dividing with the max value I'm doing because of the fact that if I don't do that I get a full white region instead of the edges of the segmented region.

Figure 1, sclera detection result:

Figure 2, result of bank of Gabor filters applied: