Find the coordinates nearest to the center coordinate of a patch in an image

i have an image of size [512 512], and xy coordinates values, which have been found on the whole image. (I have attached an example xy coordinate which i'm working on)

Now i wanted to divide the image into blocks of size [128 128], and

I need to find the center most xy coordinate that falls on each patch

[darova]

you can divide your matrix using mat2cell

• I wanted to find the center xy coordinate value that falls in each patch, and crop from the center a new patch of size [100 100]

Don't understand that. Can you explain?

[Elysi Cochin]

sir i have edited the question

Answer 1

[darova]

See this simple example

x = rand(100,1);                    % random data
y = rand(100,1);
[x1,y1] = meshgrid(0.05:.3:1);      % mesh
D = pdist2([x y],[x1(:) y1(:)]);    % combinations of distances
[~,ix] = min(D);                    % closest distances (indices)


plot(x,y,'.r')
hold on
plot(x1,y1,'.-b','markersize',25)
plot(x1',y1','.-b')
plot(x(ix),y(ix),'om','markersize',20)
hold off

Answer 2

[Rob Robinson]

centres = cell((size(ca,1)-1),(size(ca,2)-1));

for r = 1:size(ca,1)-1
    for c = 1:size(ca,2)-1 
        
        centrePoint = [(blockSizeR/2 +(blockSizeR*(r-1))) (blockSizeC/2 + (blockSizeR*(c-1)))];
        distance = sqrt(sum((xy-centrePoint).^2,2));
        [value, rowId] = min(distance);
        centres{r,c} = xy(rowId,:);
    end
end

I think this is what you meant? But this won't return any kind of error if the nearest xy point lies outside of the "patch" - something to be wary of. ( I assumed the RGB array was a matrix of values size 256x256)