Skip to content
Create a gist now

Instantly share code, notes, and snippets.

SciPy convolve2d vs Matlab conv2 'same' mode
In [198]: f = [[2,3,4],[1,6,7]]
In [199]: g = [[9,1,0],[2,5,8],[1,3,3]]
In [200]: convolve2d(f,g)
Out[200]:
array([[ 18, 29, 39, 4, 0],
[ 13, 71, 108, 51, 32],
[ 4, 26, 71, 104, 68],
[ 1, 9, 28, 39, 21]])
In [201]: convolve2d(f,g,'same')
Out[201]:
array([[ 71, 108, 51],
[ 26, 71, 104]])
In [202]: convolve2d(g,f,'same')
Out[202]:
array([[ 29, 39, 4],
[ 71, 108, 51],
[ 26, 71, 104]])
Compare with the example in http://books.google.com/books?id=fTvQkLBLyK8C&lpg=PA43&ots=0HC1LsH0cE&dq=conv2%20same%20example&pg=PA43#v=onepage&q=conv2%20same%20example&f=false
71 108 51
26 71 104
and
71 108 51
26 71 104
9 28 39
http://www.mathworks.com/help/matlab/ref/conv2.html "If there are an odd number of rows or columns, the "center" leaves one more at the beginning than the end."
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.