Skip to content
mbedded.ninja

Image Processing

Seperability

A kernel k is seperable if and only if all of it’s rows are multiples of each other. Intuitivly, this means you can pick any row l and then make a column of mulitplicative factors g, so that:

K=f⃗\crossg⃗\mathbf{K} = \vec{f} \cross \vec{g}