You are on page 1of 39

Signals Preprocessing

1. Signals contrast enhancement


2. Signals edge detection

SIGNALS EDGE DETECTION


Introduction
1. Operation of edge detection
2. Edge detection by gradient amplitude
calculation
3. Roberts operator
4. 3x3 masks edge detectors
5. Edge detection by convolution
6. Edge detection by boxcar and pyramid
operator
7. Linear Edge Crispening
8. Parallel edge detector
Conclusion
Home work
Literature

Operation of edge detection


F(j,k) 2D input signal
j=1J, k=1K
G(j,k) = f{F(j,k)} edged signal
Edge the border between signals
elements of different intensity

Operation of edge detection

Kinds of edges

Edge detection by gradient


amplitude calculation
The spatial gradient amplitude is:

2x2 pixels fragment

F(j,k)

F(j,k+1)

F(j+1,k)
F(j+1,k+1)

Edge detection by gradient


amplitude calculation
The method of discrete gradient generation

GR(j,k)= F(j,k) - F(j,k+1)


GC(j,k)= F(j,k) - F(j+1,k)

Edge detection by gradient


amplitude calculation

Edge detection by gradient


amplitude calculation

Roberts operator

2x2 pixels fragment

F(j,k)

F(j,k+1)

F(j+1,k)
F(j+1,k+1)

Roberts operator

3x3 masks edge detectors

3x3 masks edge detectors

GR(j,k)=[(A2+KA3+A4) (A0+KA7+A6)]/(K+2)
GC(j,k)=[(A0+KA1+A2) (A6+KA5+A4)] /(K+2)

3x3 masks operators


K=1 - Prewitt operator
K=2 - Sobel operator
K=21/2 - Frei-Chen operator

3x3 masks operators

Edge detection by convolution

G(j,k)= F(j,k)*H(j,k)=
F(j,k)H(j+m,k+n)
m, n = -(d-1)/2(d-1)/2
d mask size

3x3 Mask
H(j-1,k-1)
H(j-1,k)
H(j-1,k+1)

H(j,k-1)

H(j,k)
H(j,k+1)

H(j+1,k-1)
H(j+1,k)

H(j+1,k+1)

Laplassian generation of the edges

Edge detection by convolution

GR(j,k)= F(j,k)*HR(j,k)
GH(j,k)= F(j,k)*HC(j,k)

Images with noise

Edge detection by boxcar


operator

Truncated pyramid operator

Linear Edge Crispening

Linear Edge Crispening

Parallel edge detector


Roberts operator

2x2 pixels fragment

F(j,k)

F(j,k+1)

F(j+1,k)
F(j+1,k+1)

Formation of the matrixes


F(j,k)
F(j+1,k) = F(j,k) 1
F(j,k+1) = F(j,k) 1
F(j+1,k+1) = F(j,k) 11

Parallel edge detector

(Patent Nr. 845644, G06G9/00 - V.Perju


2

16
1

17

12

15

14
19

13
5
7

18

10

11

IMPULSE RESPONSE ARRAYS


FOR 3 3 EDGE
OPERATORS

Conclusion
1. Operation of edge detection
2. Edge detection by gradient amplitude
calculation
3. Roberts operator
4. 3x3 masks edge detectors
5. Edge detection by convolution
6. Laplassian generation of the edges
7. Edge detection by boxcar operator
8. Truncated pyramid operator
9. Linear Edge Crispening
10. Parallel edge detector

Home work
1. To compare different methods
2. To elaborate the algorithms for
signals edge detection by:
- gradient amplitude calculation
- Roberts operator
- Previtt edge detection
- Edge detection by convolution
- Laplass operation

Literature
S. V. Vaseghi. Advanced Digital Signal Processing and Noise
Reduction. John Willy&Sons, Ltd, 2002.
http://dsp-book.narod.ru/adsnr.htm
S.M. Kuo, B.H. Lee. Real Time Digital Signal Processing. John
Wiley&Sons, Ltd, 2001.
William K. Pratt. Digital Image Processing. John Wiley&Sons, Ltd,
2001.
. . .:, 2006
Optical Engineering Journal
www.spie.org
www.elsevier.com
www.ieee.com

You might also like