使用Gabor过滤器过滤

Question

我已经实现了Gabor滤镜，但不知道如何将其与输入图像进行卷积以获得所需的结果。我的输入图像大小为240 * 320，我将其划分为17 *的块17 ..所以总共有14 * 18 = 255个块。我也不确定Gabor滤波器的大小应该是什么？我需要将每个滤波器（14 * 18 = 252个滤波器）卷积到每个块或整个图像..

这是我的代码..

    % compute Gabor filter  
        sigma_x = 4;
        sigma_y = 4;
        w = 17;   % size of block
        wg = 5;  % size of filter
        Gabor = zeros(m1,n1);  

        for i = 9:w:240-w           
            for j3 = 9:w:320-w

                pim = im(i-ww:i+ww,j-ww:j+ww);  % block of input image
                f = F(i,j);                     %each block contains single value 
                                                of freuency and theta
                O = theta(i,j);
                [x,y] = meshgrid(-floor(wg/2):floor(wg/2));


                x_phi = x.*cos(O) + y.*sin(O);
                y_phi = -x.*sin(O) + y.*cos(O);

                x_val = ((x_phi).^2)./(sigma_x.^2);
                y_val = ((y_phi).^2)./(sigma_y.^2);

                h = exp(-0.5*(x_val+y_val)).*cos((2*pi*f).*(x_phi));
                Gabor(i-ww:i+ww,j-ww:j+ww) = conv2(pim,h,'same');

                end;              
            end; 

输出图像是指纹，不应该是黑色指纹线上的白点。