Top Banner
Vector Spaces • Space of vectors, closed under addition and scalar multiplication
41

Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Dec 19, 2015

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Vector Spaces

• Space of vectors, closed under addition and scalar multiplication

Page 2: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Image Averaging as Vector addition

Page 3: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Scaler product, dot product, norm

Page 4: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Norm of Images

Page 5: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Orthogonal Images, Distance,Basis

Page 6: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Roberts Basis: 2x2 Orthogonal

Page 7: Vector Spaces Space of vectors, closed under addition and scalar multiplication.
Page 8: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Cauchy Schwartz InequalityU+V≤U+V

Page 9: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Schwartz Inequality

Page 10: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Quotient: Angle Between two images

Page 11: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Fourier AnalysisFourier Analysis

Page 12: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Fourier Transform Pair

• Given image I(x,y), its fourier transform is

Page 13: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Image Enhancement in theFrequency Domain

Image Enhancement in theFrequency Domain

Page 14: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Complex Arithmetic

Page 15: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Fourier Traansform of an Image is a complex matrix

Let F =[F(u,v)]

F = ΦMM I(x,y) ΦNN I(x,y)= Φ*MM F Φ*MM

Where

ΦJJ (k,l)= [ΦJJ (k,l) ] and

ΦJJ (k,l) = (1/J) exp(2Πjkl/J) for k,l= 0,…,J-1

Page 16: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Fourier Transform

Page 17: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Properties

• Convolution Given the FT pair of an image

f(x,y) F(u,v) and mask pair h(x,y) H(u,v)

• f(x,y)* h(x,y) F(u,v). H(u,v) and

• f(x,y) h(x,y) F(u,v)* H(u,v)

Page 18: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Properties of Fourier TransformProperties of Fourier Transform

Page 19: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Properties of Fourier TransformProperties of Fourier Transform

Page 20: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Properties of Fourier TransformProperties of Fourier Transform

Page 21: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Properties of Fourier TransformProperties of Fourier Transform

Page 22: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Image Enhancement in theFrequency Domain

Image Enhancement in theFrequency Domain

Page 23: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Design of H(u,v)

İdeal Low Pass filter

H(u,v) = 1 if |u,v |< r

0 o.w.

Ideal High pass filter

H(u,v) = 1 if |u,v |> r

0 o.w

Ideal Band pass filter

H(u,v) = 1 if r1<|u,v |< r2

0 o.w

Page 24: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

İmage EnhancementSpatial SmoothingLow Pass Filtering

Page 25: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Ideal Low pass filterIdeal Low pass filter

Page 26: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Ideal Low Pass FilterIdeal Low Pass Filter

Page 27: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Output of the Ideal Low Pass FilterOutput of the Ideal Low Pass Filter

Page 28: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Gaussian Low Pass FilyerGaussian Low Pass Filyer

Page 29: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Gaussian Low Pass FilterGaussian Low Pass Filter

Page 30: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Gaussian Low Pass FilterGaussian Low Pass Filter

Page 31: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Gaussian Low PassFilterGaussian Low PassFilter

Page 32: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

High Pass Filter: Ideal and GaussianHigh Pass Filter: Ideal and Gaussian

Page 33: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Ideal High PassIdeal High Pass

Page 34: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Fourier Transform-High Pas Filtering

Page 35: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Frequency Spectrum of Damaged CircuitFrequency Spectrum of Damaged Circuit

Page 36: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Gaussian Low Pass and High PassGaussian Low Pass and High Pass

Page 37: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Output of Gaussian High Pass Output of Gaussian High Pass

Page 38: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Gaussian Filters: Space and Frequency DomainGaussian Filters: Space and Frequency Domain

Page 39: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Spatial Laplacian Masks and its Fourier Transform

Page 40: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Laplacian FilterLaplacian Filter

Page 41: Vector Spaces Space of vectors, closed under addition and scalar multiplication.

Laplacian FilteringLaplacian Filtering