<s>
While	O
the	O
Fourier	B-Algorithm
transform	I-Algorithm
gives	O
the	O
frequency	O
information	O
of	O
the	O
signal	O
,	O
it	O
is	O
not	O
localized	O
.	O
</s>
<s>
It	O
is	O
possible	O
to	O
use	O
a	O
short	B-Algorithm
time	I-Algorithm
Fourier	I-Algorithm
transform	I-Algorithm
for	O
this	O
purpose	O
,	O
however	O
the	O
short	B-Algorithm
time	I-Algorithm
Fourier	I-Algorithm
transform	I-Algorithm
limits	O
the	O
basis	O
functions	O
to	O
be	O
sinusoidal	O
.	O
</s>
<s>
Edge	B-Algorithm
detection	I-Algorithm
is	O
one	O
such	O
primitive	O
operation	O
,	O
where	O
the	O
edges	O
of	O
the	O
image	O
are	O
labeled	O
.	O
</s>
<s>
These	O
approaches	O
are	O
advantageous	O
because	O
unlike	O
the	O
Fourier	B-Algorithm
transform	I-Algorithm
,	O
these	O
filters	O
can	O
more	O
easily	O
represent	O
discontinuities	O
in	O
the	O
signal	O
.	O
</s>
<s>
For	O
example	O
,	O
the	O
Fourier	B-Algorithm
transform	I-Algorithm
can	O
represent	O
an	O
edge	O
,	O
but	O
only	O
by	O
using	O
an	O
infinite	O
number	O
of	O
sine	O
waves	O
.	O
</s>
<s>
For	O
reasons	O
of	O
computational	O
complexity	O
,	O
the	O
Laplacian	O
of	O
a	O
Gaussian	O
function	O
is	O
often	O
approximated	O
using	O
a	O
difference	B-Algorithm
of	I-Algorithm
Gaussians	I-Algorithm
.	O
</s>
<s>
This	O
difference	B-Algorithm
of	I-Algorithm
Gaussian	I-Algorithm
function	O
has	O
found	O
use	O
in	O
several	O
computer	O
vision	O
applications	O
such	O
as	O
keypoint	O
detection	O
.	O
</s>
<s>
In	O
practice	O
the	O
filter	O
is	O
first	O
designed	O
in	O
the	O
frequency	O
domain	O
,	O
and	O
then	O
an	O
inverse	O
Fourier	B-Algorithm
transform	I-Algorithm
gives	O
the	O
time	O
domain	O
impulse	O
response	O
.	O
</s>
