How can i extract features from sclera blood vessel structure using gabor filter.?

Hello sir,

I want to extract features from the blood vessel structure of sclera. First sclera area is detected and then blood vessels are enhanced from the sclera portion, from that enhanced blood vessel features are to be extracted.

I am trying to do feature extraction by means of GABOR FILTER... can u please tell me sir which are feature that can be extracted by means of gabor filter ??? How can i apply gabor filter to this enhanced blood vessel structured image ??

Answer 1

[Image Analyst]

There's nothing built in to MATLAB to do sclera vessel detection, so I think you should look at the published literature to find a suitable algorithm, then code it up in MATLAB. Start your search in section 20.5 here: Vision Bibliography. If you need help with MATLAB code (syntax, error messages, program flow, etc.) then come back and attach your images and m-files.

[ANJU]

Sir,

I have done sclera area detection using OTSU threshold method and after that blood vessel enhancement is done using a filter . here i am attaching the following figures that i have got.

1. input image-eye
2. sclera vessel detected using otsu
3. sclera blood vessel enhanced image

sir, can u please tell me the matlab code for extracting the features from this blood vessel enhanced image using GABOR FILTER

PLEASE HELP ME SIR...

[Asif Hasan]

I have the same problem. DId you get the solution of Gabor Filter. If so please can you give me the code that you followed. I really need that code of feature extraction. @Anju

[ANJU]

@Asif.. i didn't get the solution using gabor filter... for extracting vessel FRANGI FILTER is used in my work....

[mak carlos]

hi how can u extract the sclera part ??

[Image Analyst]

Look for white parts. I'm sure the Vision Bibliography will have some papers for describing more robust algorithms in detail.

[mak carlos]

i can not find any relate from Vision Bibliography my problem is how to detect the sclera part can u help me ??

Answer 2

[ANJU]

I used frangi filter for enhancencement and a first level wavelet transform is also used... I didnt get the output using gabor filter

[kapil khond]

@anju,i used frangi filter for sclera enhancement ,it worked thanks a lot,Now i will be doing feature extraction using Gray-Level Co-Occurrence Matrix (GLCM),as we know glcm gives the values of various features like contrast,correlation,energy ,homogeneity,entropy,cluster prominence,cluster shade and many more,please tell me from all these features which features are useful in "sclera based authentication ". also the output of frangi filter is attached please tell whether the output is appropriate or not.

Answer 3

[kapil khond]

can you tell how to detect sclera area from an eye ,thanks in advance

[Image Analyst]

It'll be the white stuff. See this for algorithms (articles): http://www.visionbib.com/bibliography/contentspeople.html#Face%20Recognition,%20Detection,%20Tracking,%20Gesture%20Recognition,%20Fingerprints,%20Biometrics

Answer 4

[kapil khond]

@anju,i am doing sclera based authentication ,i got sclera area now i want to perform sclera blood vessel enhancement , i tried that using gabor filter but instead of enhancing vessels it is enhancing eye edges ,pleae tell me what changes i need to do in code or in parametrs of gabor filter ,if you have done using different method please provide its source code thanks in advance.

below is the code of gabor filter

theta=pi/2; lambda =3.5; gamma=0.3; sigma=2.8; psi=0;

sigma_x = sigma; sigma_y = sigma/gamma;

nstds = 5; xmax = max(abs(nstds*sigma_x*cos(theta)),abs(nstds*sigma_y*sin(theta)));
 xmax = ceil(max(1,xmax));
 ymax = max(abs(nstds*sigma_x*sin(theta)),abs(nstds*sigma_y*cos(theta)));
 ymax = ceil(max(1,ymax)); 
xmin = -xmax;
 ymin = -ymax; 
[x,y] = meshgrid(xmin:xmax,ymin:ymax);

x_theta=x*cos(theta)+y*sin(theta); 
y_theta=-x*sin(theta)+y*cos(theta);

gb= exp(-.5*(x_theta.^2/sigma_x^2+y_theta.^2/sigma_y^2)).*cos(2*pi/lambda*x_theta+psi);

%subplot(3,2,5); %imshow(gb); %title('theta=...'); %imagesc(gb); %colormap(gray); %title('theta=...');

image_double=im2double(K);
 figure,subplot(2,2,1); 
%imshow(image_double);
 %figure(3);
 filtered1 = conv2(image_double,gb); 
%subplot(3,2,5); imshow(filtered1);
 title('gabor filter when theta is pi/2');
 %colormap(gray); 
%title('theta=pi/2');

Answer 5

[Mohammad Aloudat]

that is good for white sclera but what about red sclera. it share the same feature of skin.