Bab 5 Dipakai - Edge - Detection

EE465: Introduction to Digital Image
Processing Copyright Xin Li 1

DETEKSI TEPI
Dr. R. Rizal Isnanto, S.T., M.M., M.T.

Roadmap
Introduction to image analysis (computer
vision)
Its connection with psychology and neuroscience
Why is image analysis difficult?
Theory of edge detection
Gradient operator
Advanced operators
Applications
Road/sign detection in intelligent driving systems
Pupil detection in iris recognition systems
Computer Vision: the Grand Challenge
Teach a computer to see is nontrivial at all

Unlike binary images, grayscale/color images
acquired by the sensor are often easy to
understand by human being but difficult for a
machine or a robot
There are lots of interesting problems in the
field of computer vision (image analysis)
Image segmentation, image understanding, face
detection/recognition, object tracking
How does Human Vision System work?
Top-down school Bottom-up school

pixels
I see a human body
components (such as
I expect to see a human face edges, lines etc.)
I expect to see two eyes and a nose objects
Two hypothesis and nobody knows the answer yet!

An Amazing Image Example
Person A:
I see an old man with a fancy
earring and a strange hand
Person B:
I see two people on the street

and a dog lying beside
If you try really hard, you will be able
to locate at least eight different faces
from this image

Gestalt Theory (the Berlin School)
Emergence: the dog is perceived as a whole, all at once

Reification

Multistability (or Multistable Perception)

Invariance

Application: Face Detection
You are strongly encouraged to try the interactive demo out yourself
http://vasc.ri.cmu.edu/demos/faceindex/

Roadmap
Introduction to analysis of grayscale images
Why grayscale images are more difficult to
handle?
Edge detection
Gradient operator
Advanced operators
Image segmentation
Basic techniques
Texture segmentation*

Edge Detection
Why detect edge?
Edges characterize object boundaries and are
useful features for segmentation, registration
and object identification in scenes.
What is edge (to human vision system)?

No rigorous definition exists
Intuitively, edge corresponds to singularities in the image

(i.e. where pixel value experiences abrupt change)

Gradient Operators
Motivation: detect changes
change in the pixel value large gradient
image Gradient edge

Thresholding
operator map
x(m,n) g(m,n) I(m,n)
1 | g (m, n) | th
I (m, n)
0 otherwise
MATLAB function: > help edge

Common Operators
Gradient operator
g (m, n) g12 (m, n) g 22 (m, n)
Examples: 1. Roberts operator
0 1 1 0
1 0 0 1

g1 g2

Common Operators (contd)
2. Prewitt operator 3. Sobel operator
1 0 1 1 0 1
1 0 1 2 0 2
vertical
1 0 1 1 0 1
1 1 1 1 2 1
horizontal 0 0 0 0 0 0

1 1 1 1 2 1

Examples
original image horizontal edge vertical edge
Prewitt operator (th=48)

Effect of Thresholding Parameters
small threshold large

Compass Operators
1 1 0 1 1 1 0 1 1
1 0 1 0 0 0 1 0 1

0 1 1 1 1 1 1 1 0
1 0 1 1 0 1
1 0 1 1 0 1

1 0 1 1 0 1
0 1 1 1 1 1 1 1 0
1 0 1 0 0 0 1 0 1

1 1 0 1 1 1 0 1 1
g (m, n) max{| g k (m, n) |}
k
Examples
Compass operator (th=48)

Laplacian Operators
Gradient operator: first-order derivative
sensitive to abrupt change, but not slow change
2
f 2
f
second-order derivative: f 2 2
2
x y
(Laplacian operator)
2 f
0 local extreme in f
x 2
Discrete Laplacian operator

a 1 a a 0 1 0 1 1 1
1 1 4 1 1 8 1
1 a 4 1 a
1 a
a 1 a a 0 1 0 1 1 1
a=0 a=0.5
Zero Crossings
zero crossing
f f f
image Laplacian edge

zero-crossing
operator map

Examples
original image zero-crossings

Question: why is it so sensitive to noise (many false alarms)?
Answer: a sign flip from 0.01 to -0.01 is treated the same as from 100 to -100

Ideas to Improve Robustness
Linear filtering
Use a Gaussian filter to smooth out noise
component Laplacian of Gaussian
Spatially-adaptive (Nonlinear) processing
Apply different detection strategies to smooth
areas (low-variance) and non-smooth areas (high-
variance) Robust Laplacian edge detector
Return single response to edges (not multiple
edge pixels)
Hysteresis thresholding Cannys edge detector
Laplacian of Gaussian
Generalized Laplacian operator
(m 2 n 2 ) m2 n2
h(m, n) c[1 ] exp( )
2 2 2
Gaussian Laplacian edge

image
LPF () operator map
Pre-filtering: attenuate the noise sensitivity of the Laplacian

Examples
Better than Laplacian alone but still sensitive due to zero crossing

Robust Laplacian-based Edge Detector
estimate 2
local variance
Laplacian zero yes edge

image 2>th point
operator crossing?
No
No
not an
not an
edge point
edge point

Examples
More robust but return multiple edge pixels (poor localization)

Canny Edge Detector*
Low error rate of detection

Well match human perception results
Good localization of edges
The distance between actual edges in an image
and the edges found by a computational algorithm
should be minimized
Single response
The algorithm should not return multiple edges
pixels when only a single one exists

Flow-chart of Canny Edge Detector*
(J. Canny1986)
Original image
Smoothing by Gaussian convolution
Differential operators along x and y axis
Non-maximum suppression
finds peaks in the image gradient
Hysteresis thresholding locates edge strings
Edge map
Canny Edge Detector Example
original image vertical edges horizontal edges
norm of the gradient after thresholding after thinning

Marr and Hildreths Method*
Edge is scale-dependent
A different edge map can be generated at different scale
Scale space representation
f ( x , y ; s ) f ( x , y ;0 ) g ( x , y ; s )
coarse-scale fine-scale Gaussian kernel
image image with width of s
1 x2 y2
g ( x, y; s) exp( )
2s 2s
Importance of Scale

Scale-Space Edge Detection Examples
fine coarse

Image to Sketch Online Apps
http://sporkforge.com/imaging/sketch.php


Bab 5 Dipakai - Edge - Detection

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bab 5 Dipakai - Edge - Detection

Uploaded by

Copyright:

Available Formats

EE465: Introduction to Digital Image

Processing Copyright Xin Li 1

Dr. R. Rizal Isnanto, S.T., M.M., M.T.

EE465: Introduction to Digital Image

Teach a computer to see is nontrivial at all

Top-down school Bottom-up school

I expect to see two eyes and a nose objects

Two hypothesis and nobody knows the answer yet!

EE465: Introduction to Digital Image

I see two people on the street

EE465: Introduction to Digital Image

Emergence: the dog is perceived as a whole, all at once

EE465: Introduction to Digital Image

EE465: Introduction to Digital Image

EE465: Introduction to Digital Image

EE465: Introduction to Digital Image

EE465: Introduction to Digital Image

EE465: Introduction to Digital Image

What is edge (to human vision system)?

Intuitively, edge corresponds to singularities in the image

EE465: Introduction to Digital Image

change in the pixel value large gradient

image Gradient edge

MATLAB function: > help edge

g (m, n) g12 (m, n) g 22 (m, n)

Examples: 1. Roberts operator

EE465: Introduction to Digital Image

EE465: Introduction to Digital Image

original image horizontal edge vertical edge

Prewitt operator (th=48)

small threshold large

Compass operator (th=48)

Discrete Laplacian operator

image Laplacian edge

EE465: Introduction to Digital Image

original image zero-crossings

EE465: Introduction to Digital Image

Gaussian Laplacian edge

Pre-filtering: attenuate the noise sensitivity of the Laplacian

EE465: Introduction to Digital Image

EE465: Introduction to Digital Image

Laplacian zero yes edge

EE465: Introduction to Digital Image

More robust but return multiple edge pixels (poor localization)

EE465: Introduction to Digital Image

Low error rate of detection

EE465: Introduction to Digital Image

Smoothing by Gaussian convolution

Differential operators along x and y axis

Hysteresis thresholding locates edge strings

original image vertical edges horizontal edges

norm of the gradient after thresholding after thinning

A different edge map can be generated at different scale

Scale space representation

EE465: Introduction to Digital Image

EE465: Introduction to Digital Image

EE465: Introduction to Digital Image

You might also like