how to find the differnce between all the neighbouring elements in an array???

i have a 64*64 array i need to find the differnce between all the adjacent elements in the array. for instance, if i take a particular element i need to subtract it from the neighhbouring elemnts

Answer 1

[Thorsten]

If you want to add all the differences of the center pixel within it's 8-neighborhood, you can construct a filter and apply it to the image:

F = -ones(3); F(2,2) = 1;
Y = conv2(I, F)

[maria]

i dont need the sum, i would lik to get all the differences of the center pixel within it's 8-neighborhood

[Image Analyst]

Then see Matt's answer. You specify which direction you want, out of 8 possible directions. You can get them one direction at a time if that's what you want. Just be aware that convolution "flips" the window so the direction is 180 from what you might think.

[maria]

sorry its not 8 neighbourhood but 4 neighbourhood

[Image Analyst]

Same answer. Use kernels:

[0 -1 0; 0 1 0; 0 0 0]
[0 0 0; 0 1 0; 0 -1 0]
[0 0 0; -1 1 0; 0 0 0]
[0 0 0; 0 1 -1; 0 0 0]

for the 4 different directions.

[Matt J]

If you only want the 4 vertical/horizontal neighbours, it's probably easier (and faster??) to use DIFF.

Answer 2

[Matt J]

The diff() command will give you horizontal and vertical differences. Convolution can give you differences along any direction, e.g.,

 conv2(A,[0 0 0; 0 1 0; 0 0 -1]) %differences between A(i,j) and A(i-1,j-1)