[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Octave-bug-tracker] [bug #45568] imfilter incompatible to MATLAB for ev
|
From: |
Nick Jankowski |
|
Subject: |
[Octave-bug-tracker] [bug #45568] imfilter incompatible to MATLAB for even filter width |
|
Date: |
Mon, 20 Jul 2015 01:47:30 +0000 |
|
User-agent: |
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:39.0) Gecko/20100101 Firefox/39.0 |
Follow-up Comment #5, bug #45568 (project octave):
yes, sorry, didn't dig farther after seeing the conv2 in imfilter. so, with
the same imfilter call, within filter2, here's the input to conv2 in Octave
4.0.0, heading into the following line of code (after [nr,nc]=size(b))
y = conv2 (x, b(nr:-1:1, nc:-1:1), shape);
debug> x
x =
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 1 1 0 0 0 0
0 0 0 1 1 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
debug> b
b =
0 0 0 0
0 1 1 0
0 0 0 0
0 0 0 0
debug> b(nr:-1:1,nc:-1:1)
ans =
0 0 0 0
0 0 0 0
0 1 1 0
0 0 0 0
debug> shape
shape = valid
Output in octave is:
debug> conv2 (x, b(nr:-1:1, nc:-1:1), shape)
ans =
0 0 0 0 0 0
0 0 0 0 0 0
0 1 2 1 0 0
0 1 2 1 0 0
0 0 0 0 0 0
0 0 0 0 0 0
same inputs into ML conv2 give:
>> x=zeros(9,9);x(4:5,4:5)=ones(2,2)
x =
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 1 1 0 0 0 0
0 0 0 1 1 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
>> b = zeros(4,4);b([6,10])=1
b =
0 0 0 0
0 1 1 0
0 0 0 0
0 0 0 0
>> [nr,nc]=size(b)
nr =
4
nc =
4
>> shape='valid'
shape =
valid
>> conv2 (x, b(nr:-1:1, nc:-1:1), shape)
ans =
0 0 0 0 0 0
0 0 0 0 0 0
0 1 2 1 0 0
0 1 2 1 0 0
0 0 0 0 0 0
0 0 0 0 0 0
So, we get the same answer from MLs conv2 as Octave's conv2.
_______________________________________________________
Reply to this item at:
<http://savannah.gnu.org/bugs/?45568>
_______________________________________________
Message sent via/by Savannah
http://savannah.gnu.org/