Scribd
Upload
292 views

2019-21-Edge Linking and Boundary Detection

The document discusses edge linking and boundary detection techniques. It covers Laplacian edge detection, advanced edge detection including the Marr-Hildreth edge detector which uses the Laplacian of Gaussian (LoG) filter. It also discusses edge linking through local processing of edge pixel strength and direction, and region processing including polygon fitting to link edges into boundaries.

Uploaded by

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

2019-21-Edge Linking and Boundary Detection

The document discusses edge linking and boundary detection techniques. It covers Laplacian edge detection, advanced edge detection including the Marr-Hildreth edge detector which uses the Laplacian of Gaussian (LoG) filter. It also discusses edge linking through local processing of edge pixel strength and direction, and region processing including polygon fitting to link edges into boundaries.

Uploaded by

Renjeesh bkl
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 20

Contents

• Laplacian Edge Detection


• Edge Linking and Boundary Detection

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


2 7:10 PM
References

• Gonzalez and Woods, Digital Image Processing, Pearson Education,3/E, (2008)


Ch-10
The images in the presentation are taken from slides of Gonzales and Woods
and different internet websites.

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


3 7:10 PM
Laplacian Edge Detection
Laplacian of a 2D function is a second derivative:
Laplacian is defined as follows: 2
f 2
f
f 2  2
2

x y

where the partial 1st order derivatives in the x and y directions are:
2 f
 f ( x  1, y )  f ( x  1, y )  2 f ( x , y )
 x
2

2 f
 f ( x , y  1)  f ( x , y  1)  2 f ( x , y )
 y
2

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


4 7:10 PM
Laplacian Operator
Laplacian of a 2D function is a second derivative:
Not used for edge detection as
1. It is highly sensitive to noise
2.It produces double edges and is unable to detect edge direction.

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


5 7:10 PM
Advanced Edge Detection Techniques
Previous techniques does not take in to account
• image noise and
• nature of edge

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


6 7:10 PM
Marr –Hildreth Edge Detector
Marr and Hildreth suggested that Edge Detectors must have 2 salient features
• differential operator (1st or 2nd order)
• Capable of detecting blurry as well as sharply focused fine detail

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


7 7:10 PM
Marr –Hildreth Edge Detector
Operator fulfilling these conditions is a filter

• is the Laplacian operator,

• G is the 2D Gaussian function,

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


8 7:10 PM
Marr –Hildreth Edge Detector

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


9 7:10 PM
Marr –Hildreth Edge Detector

This expression is called the Laplacian of Gaussian (LoG)

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


10 7:10 PM
Laplacian of Gaussian

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


11 7:10 PM
Laplacian of Gaussian

Laplacian of Gaussian (or mexican hat) filter


• uses the gaussian for noise removal and
• Laplacian for edge detection

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


12 7:10 PM
Edge Linking and Boundary Detection
Pixels may not characterise edges completely
--due to noise,
--breaks in edges due to non uniform illumination
--other effects that introduce spurious discontinuities in intensity
values

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


13 7:10 PM
Edge Linking and Boundary Detection
1. Local processing

Two properties for establishing similarity of edge pixels


Strength of
magnitude
direction of the gradient vector

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


14 7:10 PM
Edge Linking and Boundary Detection
1.Local processing

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


15 7:10 PM
Edge Linking and Boundary Detection
2. Region
processing

a. Polygon
fitting

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


16 7:10 PM
Edge Linking and Boundary Detection
2. Region
processing

a. Polygon
fitting

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


17 7:10 PM
Edge Linking and Boundary Detection

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


18 7:10 PM
Conclusions
• LoG
• Edge Linking
Local
Region

MODULE-III EDGE LINKING AND BOUNDARY DETECTION


19 7:10 PM

You might also like