I wrote a small filter bank with 8 3x3 filters in 8 different orientations, and inside of each filter, there is a valuable v, v is decided by the magnitude of that pixel in the image to be convolved. here are my filter bank function and main program

I wrote a small filter bank with 8 3x3 filters in 8 different orientations, and inside of each filter, there is a valuable v, v is decided by the magnitude of that pixel in the image to be convolved.I want to filter the image by the 8 filters for 8 times, but the values of the each kernel will change depend on the pixel that it convolves. here are my filter bank function and main program. but when i want to call my filter bank function, matlab shows" Undefined function or variable 'v'.Error in afod (line 24)" could you please help me to find out my fault? function FB = FB(v)

FB = zeros(3,3,8);

FB(:,:,1)=[(v*v-v)/2 0 0;0 -v 0;0 0 1];

FB(:,:,2)=[0 0 (v*v-v)/2;0 -v 0;1 0 0];

FB(:,:,3)=[0 0 0;1 -v (v*v-v)/2;0 0 0];

FB(:,:,4)=[0 0 0;(v*v-v)/2 -v 1;0 0 0];

FB(:,:,5)=[1 0 0;0 -v 0;0 0 (v*v-v)/2];

FB(:,:,6)=[0 0 1;0 -v 0;(v*v-v)/2 0 0];

FB(:,:,7)=[0 1 0;0 -v 0;0 (v*v-v)/2 0];

FB(:,:,8)=[0 (v*v-v)/2 0;0 -v 0;0 1 0];

end

main program:

% read into one image

I=im2double(imread('pout.tif'));

[r,c]=size(I);

%get the minimum & maxmum gray value of image

minI=min(I(:));

maxI=max(I(:));

%expand the image

Rep=zeros(r+4,c+4);

       for x=3:3+r-1
           for y=3:3+c-1
               Rep(x,y)=I(x-3+1,y-3+1);
           end
       end

% get gradient magnitude array for this expanded image

G=imgradient(Rep);

% get the mean value of gradient magnitude of this image (find peak value)

peak=mean2(G);

% define lowest threshold a and higest threshold b

minGI=min(min(G));

maxGI=max(max(G));

a=(peak+minGI)/2;

b=(peak+maxGI)/2;

%call function FB

FB = FB(v);

for f=1:8

    h=rot90(FB(:,:,f),2);
%decide V for each pixel, and convolve adaptive kernel one pixel by one
%pixel

    for x=1:r+2

       for y=1:c+2
% get gradient magnitude for each pixel 
        gradientM=G(x+1,y+1);
        if gradientM<=a
                v=0.1;         
    elseif (a<=gradientM)&&(gradientM<=b)
        v=(((gradientM-minI)/(maxI-minI))*0.7)+0.1; 
            else
        v=0.9;
        end
        for i=1:3
            for j=1:3
                B(x,y)=B(x,y)+(Rep(x+j-1,y+i-1)*h(i,j));
            end
        end
    end
end
end
imshow(B)