Professional Documents
Culture Documents
Chapter 11:
Object Boundary
boundary vertices
(resampling)
4-directional 8-directional
chain code chain code
Partitioned boundary
Object boundary
Skeleton
2. Shape number
3. Fourier descriptor
4. Statistical moments
Shape Number
Shape number of the boundary definition: 1
the first difference of smallest magnitude
The order n of the shape number: 2 0
the number of digits in the sequence
3
Chain code:
000030032232221211
First difference:
300031033013003130
K 1
1
Fourier descriptor : a (u )
K
s
k 0
( k ) e 2uk / K
Reconstruction formula
K 1
1
s(k )
K
a
k 0
( u ) e 2uk / K
Boundary
points (Images from Rafael C. Gonzalez and Richard E.
Wood, Digital Image Processing, 2nd Edition.
Example: Fourier Descriptor
Examples of reconstruction from Fourier descriptors
1 P 1
sˆ(k ) a (u )e2uk / K
K k 0
K 1
m ri g ( ri )
where
i 0
Boundary
segment 1D graph
2. Topological Descriptors
3. Texture
4. Moments of 2D Functions
Example: Regional Descriptors
% of white pixels
Region no. compared to the
total white pixels
1 20.4%
2 64.0%
3 4.9%
4 10.7%
Infrared image of America at night (Images from Rafael C. Gonzalez and Richard E.
Wood, Digital Image Processing, 2nd Edition.
Topological Descriptors
Use to describe holes and connected components of the region
E = -1
V Q F C H E
Eg: 7-11+2=9-11= -2
The largest
After thinning
connected
area
(8479 Pixels)
(Hudson river)
B C
A
B
C
(Images from Rafael C. Gonzalez and Richard E.
Wood, Digital Image Processing, 2nd Edition.
Fourier Approach for Texture Descriptor
Concept: convert 2D spectrum into 1D graphs
Fourier
Original FFT2D coefficient Divide into areas
image +FFTSHIFT by angles
image
R0
S (q ) Sr (q )
r 1
S ( r ) Sq ( r )
Sum all pixels
in each area q 0
Fourier Approach for Texture Descriptor
Original 2D Spectrum
image (Fourier Tr.)
S(r) S(q)
5 30 312 30 12 30 12 2 321 03 2
321 03 21 03 3
30 12 21 03
2
2
6 20 02 30 12 21 03
2 2
41130 12 21 03
Example: Invariant Moments of Two-D Functions
1. Original image 2. Half size 3. Mirrored
Let x [ x1 x2 ... xn ]T
K
1
Mean: m x E{x}
K
x
k 1
k
K
1
Covariance matrix C x E{( x m x )( x m x )T }
K
k k x x
x x
k 1
T
m mT
Hotelling transformation
Let y A( x m x )
Where is a constant
6 spectral images
from an airborne
Scanner.
Component
1 3210
2 931.4
3 118.5
4 83.88
5 64.00
6 13.40
(Images from Rafael C. Gonzalez and Richard E.
Wood, Digital Image Processing, 2nd Edition.
Example: Principal Components (cont.)