calculating an expression in an efficient way

I have binary image size mxn with K number of 1's at different location (Xk,Yk). Now i want to produce a spatial voting image such that each point (1's) has the highest vote at its spatial coordinate (Xk, Yk), and its votes decrease with respect to the spatial distance. So voting matrix for X = 1.....m and Y=1.....n, can be calulated with formula shown in picuture.

http://imageshack.us/photo/my-images/31/imageqwi.png/

standard deviation sigma can be left as 1.

So how to calculate that expression in an efficient way?

Thanks

Answer 1

[Sukuchha]

this is what i tried so far.

    Image = zeros(10,10);
    % put ones
    Image(5,5)=1;
    Image(10,10)=1;
    [x,y]=meshgrid(1:10,1:10);
    %for location 5,5
    total = exp(-((x-5)^2+(y-5)^2));

But my total matrix is all zeros except total(5,5) is one. I expected to have in other cordinates some fractions. Am i doing something wrong.