Transcript Lecture 02

CS489-02 & CS589-02 Multimedia Processing
Lecture 2. Intensity Transformation and
Spatial Filtering
Spring 2009
Spatial Domain vs. Transform Domain

Spatial domain
Image plane itself, directly process the intensity
values of the image plane

Transform domain
Process the transform coefficients, not directly
process the intensity values of the image plane
4/13/2015
2
Spatial Domain Process
g ( x, y )  T [ f ( x, y )])
f ( x, y ) : input image
g ( x, y ) : output image
T : an operator on f defined over
a neighborhood of point ( x, y)
4/13/2015
3
Spatial Domain Process
4/13/2015
4
Spatial Domain Process
Intensity transformation function
s  T (r )
4/13/2015
5
Some Basic Intensity Transformation Functions
4/13/2015
6
Image Negatives
Image negatives
s  L 1  r
4/13/2015
7
Example: Image Negatives
Small
lesion
4/13/2015
8
Log Transformations
Log Transformations
s  c log(1  r )
4/13/2015
9
Example: Log Transformations
4/13/2015
10
Power-Law (Gamma) Transformations
s  cr
4/13/2015

11
Example: Gamma Transformations
4/13/2015
12
Example: Gamma Transformations
4/13/2015
13
Piecewise-Linear Transformations

Contrast Stretching
— Expands the range of intensity levels in an image
― spans the full intensity range of the recording medium

Intensity-level Slicing
— Highlights a specific range of intensities in an image
4/13/2015
14
4/13/2015
15
Highlight the major
blood vessels and
study the shape of the
flow of the contrast
medium (to detect
blockages, etc.)
4/13/2015
Measuring the actual
flow of the contrast
medium as a function
of time in a series of
images
16
Bit-plane Slicing
4/13/2015
17
Example
4/13/2015
18
Example
4/13/2015
19
Histogram Processing

Histogram Equalization

Histogram Matching

Local Histogram Processing

Using Histogram Statistics for Image
Enhancement
4/13/2015
20
Histogram Processing
Histogram h( rk )  nk
rk is the k th intensity value
nk is the number of pixels in the image with intensity rk
nk
Normalized histogram p (rk ) 
MN
nk : the number of pixels in the image of
size M  N with intensity rk
4/13/2015
21
4/13/2015
22
Histogram Equalization
The intensity levels in an image may be viewed as
random variables in the interval [0, L-1].
Let pr (r ) and ps ( s) denote the probability density
function (PDF) of random variables r and s.
4/13/2015
23
Histogram Equalization
s  T (r )
0  r  L 1
a. T(r) is a strictly monotonically increasing function
in the interval 0  r  L -1;
b. 0  T (r )  L -1 for 0  r  L -1.
4/13/2015
24
Histogram Equalization
s  T (r )
0  r  L 1
a. T(r) is a strictly monotonically increasing function
in the interval 0  r  L -1;
b. 0  T (r )  L -1 for 0  r  L -1.
T (r ) is continuous and differentiable.
ps (s)ds  pr (r )dr
4/13/2015
25
Histogram Equalization
r
s  T (r )  ( L  1)  pr (w)dw
0
ds dT (r )
d  r


 ( L  1)
p
(
w
)
dw
r



0
dr
dr
dr 
 ( L 1) pr (r )
pr (r )dr pr (r )
1
pr ( r )
ps ( s ) 



 ( L  1) pr (r )  L  1
 ds 
ds
 
 dr 
4/13/2015
26
Example
Suppose that the (continuous) intensity values
in an image have the PDF
 2r
,

2
pr (r )   ( L  1)
 0,

for 0  r  L-1
otherwise
Find the transformation function for equalizing
the image histogram.
4/13/2015
27
Example
r
s  T (r )  ( L  1)  pr (w)dw
0
 ( L  1) 
r
0
2w
dw
2
( L  1)
2
r

L 1
4/13/2015
28
Histogram Equalization
Continuous case:
r
s  T (r )  ( L  1)  pr (w)dw
0
Discrete values:
k
sk  T (rk )  ( L  1) pr (rj )
j 0
L 1 k
 ( L  1)

nj

MN j 0
j 0 MN
k
4/13/2015
nj
k=0,1,..., L-1
29
Example: Histogram Equalization
Suppose that a 3-bit image (L=8) of size 64 × 64 pixels (MN = 4096)
has the intensity distribution shown in following table.
Get the histogram equalization transformation function and give the ps(sk)
for each sk.
4/13/2015
30
Example: Histogram Equalization
0
s0  T (r0 )  7 pr (rj )  7  0.19  1.33
1
s1  T (r1 )  7 pr (rj )  7  (0.19  0.25)  3.08
3
j 0
1
j 0
4/13/2015
s2  4.55  5
s3  5.67  6
s4  6.23  6
s5  6.65  7
s6  6.86  7
s7  7.00  7
31
Example: Histogram Equalization
4/13/2015
32
4/13/2015
33
4/13/2015
34
Question
Is histogram equalization always good?
No
4/13/2015
35
Histogram Matching
Histogram matching (histogram specification)
—A processed image has a specified histogram
Let pr (r ) and pz ( z ) denote the continous probability
density functions of the variables r and z. pz ( z ) is the
specified probability density function.
Let s be the random variable with the probability
r
s  T (r )  ( L  1)  pr ( w)dw
0
Obtain a transformation function G
z
G ( z )  ( L  1)  pz (t )dt  s
0
4/13/2015
36
Histogram Matching
r
s  T (r )  ( L  1)  pr ( w)dw
0
z
G ( z )  ( L  1)  pz (t ) dt  s
0
1
z  G (s)  G
4/13/2015
1
T (r)
37
Histogram Matching: Procedure

Obtain pr(r) from the input image and then obtain the
values of s
r
s  ( L  1)  pr ( w)dw
0

Use the specified PDF and obtain the transformation
function G(z)
z
G( z )  ( L  1) pz (t )dt  s
0

Mapping from s to z
z  G 1 (s)
4/13/2015
38
Histogram Matching: Example
Assuming continuous intensity values, suppose that an
image has the intensity PDF
 2r
,

2
pr (r )   ( L  1)
 0,

for 0  r  L -1
otherwise
Find the transformation function that will produce an
image whose intensity PDF is
 3z 2
,

3
pz ( z )   ( L  1)
 0,

4/13/2015
for 0  z  ( L -1)
otherwise
39
Histogram Matching: Example
Find the histogram equalization transformation for the input
image
r
r
0
0
s  T (r )  ( L  1)  pr ( w)dw  ( L  1) 
2
r
2w
dw 
2
( L  1)
L 1
Find the histogram equalization transformation for the specified
histogram
z
z
0
0
G( z )  ( L  1)  pz (t )dt  ( L  1) 
The transformation function
1/3
z  ( L  1) s 
2
4/13/2015
3t 2
z3
dt 
s
3
2
( L  1)
( L  1)
1/3


2 r
 ( L  1)

L

1


2
2 1/3
 ( L  1)r 
40
Histogram Matching: Discrete Cases

Obtain pr(rj) from the input image and then obtain the
values of sk, round the value to the integer range [0, L1].
k
k
( L  1)
sk  T (rk )  ( L  1) pr (rj ) 
nj

MN j 0
j 0

Use the specified PDF and obtain the transformation
function G(zq), round the value to the integer range [0,
L-1].
q
G( zq )  ( L  1) pz ( zi )  sk
i 0

Mapping from sk to zq
zq  G1 (sk )
4/13/2015
41
Example: Histogram Matching
Suppose that a 3-bit image (L=8) of size 64 × 64 pixels (MN = 4096)
has the intensity distribution shown in the following table (on the
left). Get the histogram transformation function and make the output
image with the specified histogram, listed in the table on the right.
4/13/2015
42
Example: Histogram Matching
Obtain the scaled histogram-equalized values,
s0  1, s1  3, s2  5, s3  6, s4  7,
s5  7, s6  7, s7  7.
Compute all the values of the transformation function G,
0
G( z0 )  7 pz ( z j )  0.00  0
j 0
G ( z1 )  0.00
G ( z3 )  1.05
0
1
G ( z5 )  4.55  5
G ( z7 )  7.00  7
4/13/2015
G( z2 )  0.00  0
G( z4 )  2.45  2
G( z6 )  5.95  6
43
Example: Histogram Matching
4/13/2015
44
Example: Histogram Matching
Obtain the scaled histogram-equalized values,
s0  1, s1  3, s2  5, s3  6, s4  7,
s5  7, s6  7, s7  7.
Compute all the values of the transformation function G,
0
G( z0 )  7 pz ( z j )  0.00  0
j 0
G ( z1 )  0.00
G ( z3 )  1.05
0
 1 s0
G ( z5 )  4.55  5
G ( z7 )  7.00  7
4/13/2015
s2
G( z2 )  0.00  0
G( z4 )  2.45  2 s1
G( z6 )  5.95  6 s3
s4 s5 s6 s7
45
Example: Histogram Matching
s0  1, s1  3, s2  5, s3  6, s4  7,
s5  7, s6  7, s7  7.
rk
4/13/2015
0
1
2
3
4
5
6
7
46
Example: Histogram Matching
rk  zq
03
1 4
25
36
47
57
67
77
4/13/2015
47
Example: Histogram Matching
4/13/2015
48
Example: Histogram Matching
4/13/2015
49
4/13/2015
50
Local Histogram Processing
Define a neighborhood and move its center from pixel to
pixel
At each location, the histogram of the points in the
neighborhood is computed. Either histogram equalization or
histogram specification transformation function is obtained
Map the intensity of the pixel centered in the neighborhood
Move to the next location and repeat the procedure
4/13/2015
51
Local Histogram Processing: Example
4/13/2015
52
Using Histogram Statistics for Image
Enhancement
Average Intensity
L 1
m   ri p(ri )
i 0
1

MN
M 1 N 1
 f ( x, y)
x 0 y 0
L 1
un (r )   (ri  m)n p(ri )
i 0
Variance
L 1
  u2 (r )   (ri
2
i 0
4/13/2015
1
2

 m) p(ri )
MN
M 1 N 1
 f ( x, y)  m
2
x 0 y 0
53
Using Histogram Statistics for Image
Enhancement
Local average intensity
L 1
msxy   ri psxy (ri )
i 0
sxy denotes a neighborhood
Local variance

4/13/2015
L 1
2
sxy
  (ri  msxy ) psxy (ri )
2
i 0
54
Using Histogram Statistics for Image
Enhancement: Example
 E f ( x, y ), if msxy  k0 mG and k1 G   sxy  k2 G
g ( x, y )  
otherwise
 f ( x, y ),
mG : global mean;
 G : global standard deviation
k0  0.4; k1  0.02; k2  0.4; E  4
4/13/2015
55
Spatial Filtering
A spatial filter consists of (a) a neighborhood, and (b) a
predefined operation
Linear spatial filtering of an image of size MxN with a filter
of size mxn is given by the expression
g ( x, y) 
a
b
  w(s, t ) f ( x  s, y  t )
s  a t  b
m  2a  1;
4/13/2015
n  2b  1
56
Spatial Filtering
4/13/2015
57
Spatial Correlation
The correlation of a filter w( x, y) of size m  n
with an image f ( x, y), denoted as w( x, y ) f ( x, y)
w( x, y)
f ( x, y) 
a
b
  w(s, t ) f ( x  s, y  t )
s  a t  b
4/13/2015
58
Spatial Convolution
The convolution of a filter w( x, y ) of size m  n
with an image f ( x, y), denoted as w( x, y ) f ( x, y)
w( x, y)
f ( x, y) 
a
b
  w(s, t ) f ( x  s, y  t )
s  a t  b
4/13/2015
59
4/13/2015
60
Smoothing Spatial Filters
Smoothing filters are used for blurring and for noise
reduction
Blurring is used in removal of small details and bridging of
small gaps in lines or curves
Smoothing spatial filters include linear filters and nonlinear
filters.
4/13/2015
61
Spatial Smoothing Linear Filters
The general implementation for filtering an M  N image
with a weighted averaging filter of size m  n is given
a
g ( x, y ) 
b
  w(s, t ) f ( x  s, y  t )
s  a t  b
a
b
  w(s, t )
s  a t  b
where m  2a  1,
4/13/2015
n  2b  1.
62
Two Smoothing Averaging Filter Masks
4/13/2015
63
4/13/2015
64
Example: Gross Representation of Objects
4/13/2015
65
Order-statistic (Nonlinear) Filters
— Nonlinear
— Based on ordering (ranking) the pixels contained in the
filter mask
— Replacing the value of the center pixel with the value
determined by the ranking result
E.g., median filter, max filter, min filter
4/13/2015
66
Example: Use of Median Filtering for Noise Reduction
4/13/2015
67
Sharpening Spatial Filters
►
Foundation
►
Laplacian Operator
►
Unsharp Masking and Highboost Filtering
►
Using First-Order Derivatives for Nonlinear Image
Sharpening
4/13/2015
68
Sharpening Spatial Filters: Foundation
►
The first-order derivative of a one-dimensional function f(x)
is the difference
f
 f ( x  1)  f ( x)
x
►
The second-order derivative of f(x) as the difference
2 f
 f ( x  1)  f ( x  1)  2 f ( x)
2
x
4/13/2015
69
4/13/2015
70
Sharpening Spatial Filters: Laplace Operator
The second-order isotropic derivative operator is the
Laplacian for a function (image) f(x,y)
2
2

f

f
2
 f  2  2
x
y
2 f
 f ( x  1, y )  f ( x  1, y)  2 f ( x, y)
2
x
2 f
 f ( x, y  1)  f ( x, y  1)  2 f ( x, y)
2
y
2 f  f ( x  1, y)  f ( x  1, y )  f ( x, y  1)  f ( x, y  1)
- 4 f ( x, y)
4/13/2015
71
Sharpening Spatial Filters: Laplace Operator
4/13/2015
72
Sharpening Spatial Filters: Laplace Operator
Image sharpening in the way of using the Laplacian:
g ( x, y)  f ( x, y)  c  2 f ( x, y) 
where,
f ( x, y ) is input image,
g ( x, y ) is sharpenend images,
c  -1 if  2 f ( x, y ) corresponding to Fig. 3.37(a) or (b)
and c  1 if either of the other two filters is used.
4/13/2015
73
4/13/2015
74
Unsharp Masking and Highboost Filtering
►
Unsharp masking
Sharpen images consists of subtracting an unsharp (smoothed)
version of an image from the original image
e.g., printing and publishing industry
►
Steps
1. Blur the original image
2. Subtract the blurred image from the original
3. Add the mask to the original
4/13/2015
75
Unsharp Masking and Highboost Filtering
Let f ( x, y ) denote the blurred image, unsharp masking is
g mask ( x, y )  f ( x, y )  f ( x, y )
Then add a weighted portion of the mask back to the original
g ( x, y )  f ( x, y )  k * g mask ( x, y )
k 0
when k  1, the process is referred to as highboost filtering.
4/13/2015
76
Unsharp Masking: Demo
4/13/2015
77
Unsharp Masking and Highboost Filtering: Example
4/13/2015
78
Image Sharpening based on First-Order Derivatives
For function f ( x, y ), the gradient of f at coordinates ( x, y )
is defined as
 f 
 g x   x 
f  grad( f )      
 g y   f 
 y 
The magnitude of vector f , denoted as M ( x, y)
Gradient Image
4/13/2015
M ( x, y)  mag(f )  g x 2  g y 2
79
Image Sharpening based on First-Order Derivatives
The magnitude of vector f , denoted as M ( x, y)
M ( x, y)  mag(f )  g x 2  g y 2
M ( x, y) | g x |  | g y |
z1
z4
z7
4/13/2015
z2
z5
z8
z3
z6
z9
M ( x, y) | z8  z5 |  | z6  z5 |
80
Image Sharpening based on First-Order Derivatives
Roberts Cross-gradient Operators
M ( x, y) | z9  z5 |  | z8  z6 |
z1
z4
z7
4/13/2015
z2
z5
z8
z3
z6
z9
Sobel Operators
M ( x, y) | ( z7  2 z8  z9 )  ( z1  2 z2  z3 ) |
 | ( z3  2 z6  z9 )  ( z1  2 z4  z7 ) |
81
Image Sharpening based on First-Order Derivatives
4/13/2015
82
Example
4/13/2015
83
Example:
Combining Spatial
Enhancement
Methods
Goal:
Enhance the
image by
sharpening it and
by bringing out
more of the
skeletal detail
4/13/2015
84
Example:
Combining Spatial
Enhancement
Methods
Goal:
Enhance the
image by
sharpening it and
by bringing out
more of the
skeletal detail
4/13/2015
85