Advanced Digital Signal Processing - PowerPoint PPT Presentation

1 / 43
About This Presentation
Title:

Advanced Digital Signal Processing

Description:

Short-term Fourier Transform. Two test signals again. x(t)=cos( 1 t) cos( 2t) cos( 3) cos ... Problem: the same and t throught the entire plane! ... – PowerPoint PPT presentation

Number of Views:603
Avg rating:3.0/5.0
Slides: 44
Provided by: Jus6
Category:

less

Transcript and Presenter's Notes

Title: Advanced Digital Signal Processing


1
DEPARTMENT OF COMPUTER SCIENCE UNIVERSITY OF
JOENSUU JOENSUU, FINLAND
  • Advanced Digital Signal Processing
  • Lecture 16
  • Wavelet Transform
  • Alexander Kolesnikov
  • (16.11.2005)

2
Short-term Fourier Transform
3
Two test signals again
a)
?1 10? ?2 20? ?3 40? ?4100?
x(t)cos(?1 t)cos(?2t)cos(?3)cos(?4t)
b)
x1(t)cos(?1t) x2(t)cos(?2t) x3(t)cos(?3t) x4(t)
cos(?4t)
x1(t) x2(t) x3(t) x4(t)
4
Spectrums of the test signals
Signals are different, spectrums are similar
Why?
5
What is wrong with the Fourier Transform?
Two basis functions sin(?t) and ?(t)
Support region In time In frequency
sin(?t) ? 0
?(t) 0 ? The basis function sin(?t) is
not localized in time, The ?(t) (sample) is not
localized in frequency.
Fourier Transform is good for stationary
signals, but real signals are non-stationary
ones. Another problem signals with
discontinuties.
6
Short-term Fourier Transform (STFT)
signal x(t)
window h(t-?)
signal in window
7
Short-term Fourier Transform (STFT)
Introduce basis functions which are compact
in time and frequency domains. In other words,
let us divide the input signal into
time sub-intervals, and perform DFT for every
sub-interval
8
STFT Time-frequency plane
Equidistant frequencies
??
?t
9
Problems with STFT
Uncertainty Principle
We cannot localize events in time and frequency
simultaneously!
Improved time resolution?Degraded frequency
resolution Improved frequency resolution
?Degraded time resolution
Problem the same ?? and ?t throught the entire
plane!
10
Wavelet Transform
11
The main idea
Keep the relative accuracy in the frequency
domain
Use wider time window for lower frequencies!
12
Time-frequency plane
13
Scale Illustration
Coast line
  • Coarse scale
  • flying in a jet at 5 km
  • Medium scale
  • bird flight at 100 m
  • Fine scale
  • beach walk

14
Wavelet transform wavelet mother function
Two properties of mother wavelet function
mother wavelet
baby wavelets
15
Wavelet transform wavelet mother function
How to obtain a set of wavelet functions?
Translation (?) and dilation (scaling, s)
16
Scaling (stretching or compressing)
s1
s0.5
s0.25
17
Translation (shift)
18
Translation and stretching
s2-7
s2-6
s2-6
s2-6
s2-5
s2-4
s2-3
s2-3
19
Localization in time
Baby wavelets Mother wavelet
...
...
k0,1,...,
20
Examples of mother wavelets
21
Haar wavelets
Scaling function
22
Discrete wavelet transform
j1
j
k
23
Hierarchical calculation of DWT
24
Haar wavelet transform
25
Haar DWT Example
Input data Xx1,x2,x3,, x16 Haar wavelet
transform (a,b)?(s,d) where 1) scaling
function s(ab)/2 (smooth, LPF) 2) Haar
wavelet d(a-b) (details, HPF)
X10,13, 11,14, 12,15, 12,14, 12,13, 11,13,
10,11 11.5,12.5, 13.5,13, 12.5,12, 10.5
-3, -3, -3, -2, -1,-2,-1 12, 13.25,
12.25, 10.5 -2,0.5,-0.5 -3, -3, -3, -2,
-1,-2,-1 12.625, 11.375 -1.25, 1.75
-2,0.5,-0.5 -3, -3, -3, -2, -1,-2,-1
121.25 -1.25, 1.75 -2,0.5,-0.5 -3, -3,
-3, -2, -1,-2,-1
26
Inverse Haar DWT Example
Inverse Haar wavelet transform (s,d)? (a,b)
1) asd/2 2) bs?d/2
X10,13, 11,14, 12,15, 12,14, 12,13, 11,13,
10,11? 11.5,12.5, 13.5,13, 12.5,12, 10.5
-3, -3, -3, -2, -1,-2,-1 12, 13.25,
12.25, 10.5 -2,0.5,-0.5 -3, -3, -3, -2,
-1,-2,-1 12.625, 11.375 -1.25, 1.75
-2,0.5,-0.5 -3, -3, -3, -2, -1,-2,-1
121.25 -1.25, 1.75 -2,0.5,-0.5 -3, -3,
-3, -2, -1,-2,-1
Y 121.25 -1.25, 1.75 -2,0.5,-0.5 -3,
-3, -3, -2, -1,-2,-1 12.625,11.375
-1.25, 1.75 -2,0.5,-0.5 -3, -3, -3, -2,
-1,-2,-1 12, 13.25, 12.25, 10.5
-2,0.5,-0.5 -3, -3, -3, -2, -1,-2,-1
11.5,12.5, 13.5,13, 12.5,12, 10.5 -3, -3,
-3, -2, -1,-2,-1 10,13, 11,14, 12,15,
12,14, 12,13, 11,13, 10,11
27
2-D Wavelet transform
28
2-D wavelet transform
Transform Coeff. 4123, -12.4, -96.7, 4.5,
Original 128, 129, 125, 64, 65,
29
Wavelet transform as Subband Transform
30
Wavelet Transform and Filter Banks
31
Wavelet Transform and Filter Banks
h0(n) is scaling function, low pass filter
(LPF) h1(n) is wavelet function, high pass filter
(HPF)
is subsampling (decimation)
32
5/3 filter for lossless encoding (JPEG2000)
33
9/7 filter for lossy encoding (JPEG2000)
34
5-tab low pass filter (LPF)
2
Filtration
Subsampling
35
3-tab high pass filter (HPF)
2
Subsampling
Filtration
36
Filtration with 5/3-filters
Input x
Input x
LPF s
HPF d
37
Inverse wavelet transform
is up-sampling (zeroes inserting)
Synthesis filters g0(n)?(-1)nh1(n)
g1(n)?(-1)nh0(n)
38
Complexity of discrete wavelet transform
Without scaling-function property
39
Scaling function and Wavelets
Scaling function
Wavelet function
where
This is very important property of the
wavelets. To obtain WT coefficients for level j
we can process WT coefficients for level j1.
? DWT Complexity O(N)
40
Haar Scaling function and Wavelets
41
Daubechies wavelets of order 2
?(t)
?(t)
Scaling function Wavelet
function
42
Multiresolution discrete wavelet analysis
Low-resolution approximation
Wavelets details
43
Applications
1. Image compression (JPEG200) 2. Signal
filtering (denoising) 3. Signal analysis 4. etc.
Write a Comment
User Comments (0)
About PowerShow.com