Wavelets and Image Processing, why bother - PowerPoint PPT Presentation

1 / 14
About This Presentation
Title:

Wavelets and Image Processing, why bother

Description:

'These turn out, for about 97% of the neurons, to be very closely ... JPEG at that compression rate was 'blocky' Used wavelet transform/scalar quantization (WSQ) ... – PowerPoint PPT presentation

Number of Views:71
Avg rating:3.0/5.0
Slides: 15
Provided by: cns6
Category:

less

Transcript and Presenter's Notes

Title: Wavelets and Image Processing, why bother


1
Wavelets and Image Processing, why bother?
  • Cn780
  • 5/4/04
  • Tim McKenna

2
Outline
  • Multiresolution Analysis using Wavelets
  • Compression
  • Wavelets for a Vision
  • The FBI Fingerprint Database
  • Space-variant object recognition using wavelets

3
Multiresolution Analysis Using Wavelets
  • The claim by R.G.Daugman concerning the receptive
    field profile of neurons in visual cortex
  • - These turn out, for about 97 of the neurons,
    to be very closely described as 2D Gabor wavelets

4
Gabor Wavelets
Functional Form
The Children of the Mother
Mother
where
5
Compression
Lai et al 1999, Yale database
  • Compression in a Nutshell
  • DWT the image -gtwavelet coefficients
  • throw out the small coefficients

6
Foveated Vision
  • Space-variant wavelet representation with a
    log-polar flavor.

7
  • Orthonormal Wavelet Bases (right) (scaling
    functions on left)
  • Meyer
  • Battle-Lemarie
  • Battle-Lamarie
  • Haar
  • Daubechies
  • other

Daubechies 1992
8
Wavelets for a Vision
Chang,MallatYap 1999
Three Mother Wavelets
9
from 1 dimension to 2
One-dimension foveation operator
Generalization to two dimensions
10
Chang,MallatYap 1999
11
The FBI Fingerprint Database
  • Single print 1/2megabyte, (768x768)
  • Target rate desired .75bits/pixel
  • JPEG at that compression rate was blocky
  • Used wavelet transform/scalar quantization (WSQ)

12
WSQ
http//www.c3.lanl.gov/brislawn/FBI/OverCompress/
HowItWorks/howitworks.html
13
  • Discrete Wavelet Transform (DWT) Decomposition
  • 64 subbands
  • Variance varies

14
Space-variant object recognition using wavelets
Write a Comment
User Comments (0)
About PowerShow.com