Image Processing Reference
In-Depth Information
Try these out
!
We might also want to use a more specialised form of look-up table, say the
saw-tooth
operator. For this, we split the brightness range up into bands, and use a linear look-up
table in each.
saw_tooth
bright
:=mod(bright,60)
and use the modulus operator to give a saw_tooth
function
50
saw
_tooth
bright
0
0
100
200
Bright
So we'll define a saw-tooth function as:
saw_tooth(brightness,factor):=mod(brightness,factor)
And as a function it is
for x
∈
0..cols(pic)-1
Address the whole picture
saw(pic,modulus):=
for y
∈
0..rows(pic)-1
newpic
y,x
←
Apply saw_tooth
saw_tooth
(pic
y,x
,
modulus)
Output the picture
newpic
So let's saw it:
sawn:=saw(eye,60)
A common use of point functions is to
equalise
the intensity response of a camera. We
work out the
histogram
of the camera response. This gives a function which can equalise
the combined response of function*camera equal to
unity
, to give a constant intensity
response. Let us suggest that the known performance of the camera is
exponential
. The
equalising function is
logarithmic
since log(exp(q))=q. So let's see what it's like:
