crop circle region from image

hi, im currently doing a iris recognition. im now in a segmentation process. ive already detect the the pupil region but how can i remove the background and leave only the pupil region?

[AKHILA P]

how you do the pupil region detection? please help me i am also doing iris recognition

Answer 1

[KSSV]

This code shall work for you. When prompted/ figure opens, you need two click twice.

1. First click, click at the center of the circle, where you want to draw a circle.

2. Second click, click at any point which is at a distance of radius of circle you wanted to draw.

I = imread('Capture111.jpg');
[nx,ny,d] = size(I) ;
[X,Y] = meshgrid(1:ny,1:nx) ;
imshow(I) ;
hold on
[px,py] = getpts ;   % click at the center and approximate Radius

r = sqrt(diff(px).^2+diff(py).^2) ;
th = linspace(0,2*pi) ;

xc = px(1)+r*cos(th) ; 
yc = py(1)+r*sin(th) ; 
plot(xc,yc,'r') ;
% Keep only points lying inside circle
idx = inpolygon(X(:),Y(:),xc',yc) ;

for i = 1:d
    I1 = I(:,:,i) ;
    I1(~idx) = 255 ;
    I(:,:,i) = I1 ;
end

figure
imshow(I)

I am attaching the result, I got on running the above code:

[KSSV]

Then how do you think, I got the attached image?

[Matthew Burges]

how to do this without selecting the center and radius, meaning matlab should do this step on its own.

[Sruthi Radhakrishnan]

Thank you for the code, But how to eliminate the white background in the image ?

[Gustavo Deza]

If I multiple circle I want to extract, and have a variable with the values of the centers and radii, how can I make a loop to extract all the circles?

for i_1 = 1:size(centers)   
    px = centers(i_1,1);
    py = centers(i_1,2);
    r = radii(i_1) ;
    th = linspace(0,2*pi) ;
    xc = px(i_1)+r(i_1)*cos(th) ;
    yc = py(i_1)+r(i_1)*sin(th) ;
    %plot(xc,yc,'r') ;
    % Keep only points lying inside circle
    idx = inpolygon(X(:),Y(:),xc',yc) ;
    for j = 1:d
        I1 = f_gray(:,:,j) ;
        I1(~idx) = 255 ;
        I(:,:,j) = I1 ;
    end
end

I get this error: "Index exceeds the number of array elements (1)."