Scribd
Upload
48 views

Correlation & Convolution

Uploaded by

Keilla Romabiles Leopando
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
48 views

Correlation & Convolution

Uploaded by

Keilla Romabiles Leopando
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 17

( CONTINUOUS – TIME SIGNAL)

an operation used to measure the


similarity between two signals or
functions and the time relation of
the similarity

Correlation Function

f12   f1( ) f 2 (t   )d

 to correlate a f1(t )
signal
 with another f 2 (t )
signal
 first, shift the f 2 (t   )
second signal
 then multiply f1( ) f 2 (t   )
both signals

 then integrate
under the curve 

f1( ) f 2 (t   )d
 Autocorrelation
 if the signal f1 and f2 are the same
 provides information about the time – domain
structure of a noisy signal
 often used to discover periodic components in noisy
signals
 Cross Correlation
 if the signal f1 and f2 are different
 used to detect and locate known reference signal
buried in noise of known reference signals
 used to identify a signal by comparison with a
library
 signal processing
 pattern recognition
 single particle analysis
 electron tomographic averaging
 cryptanalysis
 neurophysiology
 N=1024; % Number of  subplot(3,1,2);
samples to generate  plot(y);
 f=1; % Frequency of the  title('y(n), Pure Sinewave +
sinewave Noise');
 FS=200; % Sampling  grid;
frequency
 Rxy=xcorr(x,y); % Estimate
 n=0:N-1; % Sampling index the cross correlation
 x=sin(2*pi*f1*n/FS); %  subplot(3,1,3);
Generate x(n)
 plot(Rxy);
 y=x+10*randn(1,N); %
Generate y(n)  title('Cross correlation Rxy');
 subplot(3,1,1);  grid;
 plot(x);
 title('Pure Sinewave');
 grid;

Source: Digital Signal Processing Case Study Copyright-S.R.Taghizadeh


Pure Sinewave
1

-1
0 200 400 600 800 1000 1200
y(n), Pure Sinewave + Noise
50

-50
0 200 400 600 800 1000 1200
Cross correlation Rxy
1000

-1000
0 500 1000 1500 2000 2500
a mathematical operation on two
functions f and g, producing a
third function that is typically
viewed as a modified version of
one of the original functions
 used in communications similar to
cross – correlation except that the
second signal is flipped back to
front

f12 (t )   f1( ) f 2 (t   )d


•the convolution formula can be described as


a weighted sum of the function ƒ1() at the
moment t where the weighting is given by
f2(−) simply shifted by amount t
• as t changes, the weighting function
emphasizes different parts of the input
function
x(t )  h(t )  h(t )  x(t )
x(t )  [h(t )  z (t )]  [ x(t )  h(t )]  z (t )
x(t )  [h(t )  z (t )]  x(t )  h(t )  x(t )  z (t )
x(t )  [ch(t )]  c[ x(t ) * h(t )]
d  dh(t ) 
[ x(t ) * h(t )]  x(t ) *  
dt  dt 
v(t ) * w(t )  V ( f ) W ( f )

v(t ) * w(t )  V ( s )  W ( s )

v(t )  w(t )  V ( f ) *W ( f )
 to convolve a f1(t )
signal
 with another f 2 ( )
signal
 first, flip the
f 2 ( )
second signal
f 2 (t   )
 then, shift it
 then multiply
both signals f1( ) f 2 (t   )
 then integrate

under the curve 

f1( ) f 2 (t   )d
 applications
 probability
 statistics
 computer vision
 image and signal processing (e.g. design
and implementation of finite impulse
response filters)
 electrical engineering
 differential equations
 telescope
 microscope
vt  wt 
v(t )  t 0t T
w(t )  sin( t ) 0  t  

vt  wt 
v(t )  t 0t T
w(t )  2e t
0t 
autocorrelation
w(t )  2e 2t
0t 
autocorrelation
w(t )  sin t  0t 

vt  wt 
v(t )  t 0t T
w(t )  2 cost  0t 

You might also like