Chapter 3

Image Enhancement in the Spatial

Domain
 Background
Spatial domain refers to the aggregated
of pixels composing an image.
Spatial domain methods are procedures
that operate directly on these pixels.
Spatial domain processes will be denoted
by the expression
g(x,y)=T [ f(x,y) ]
where f(x,y) is the input image, g(x,y) is the processed image,
and T is an operator of f 2
 Some Basic Gray Level
Transformations
Discussing gray-level transformation
functions. The value of pixels, before and
after processing are related by an
expression of the form
s = T(r)
Where r is values of pixel before process
s is values of pixel after process
T is a transformation that maps a pixel
value r into a pixel value s. 3
 Gray Level Transformations
Image Negatives
Log Transformations
Exponential Transformations
Power-Law Transformations

4
 Image Negatives
The negative of an image with gray L
levels is given by the expression
s=L–1–r
Where r is value of input pixel, and
s is value of processed pixel

5
6
 Log Transformations
The general form of the log
transformation shown in
s = c log(1+r)

Where c is constant, and it is assumed that r0

7
C = 1.0

C = 0.8

8
 Exponential Transformations
The general form of the log
transformation shown in
s = c exp(r)

Where c is constant, and it is assumed that r0

9
C = 1.0

C = 0.8

10
 Power-Law Transformations
Power-law transformations have the
basic form

s  cr
Where c and  are positive constant.
Sometime above Equation is written as

s  c(r   )

11
  = 0.5

 = 1.0

 = 5.0

12
 Piecewise-Linear
Transformation Functions
Translate Mapping
Contrast stretching
Gray-level slicing
Bit-plane slicing

13
Translate Mapping

14
 Contrast stretching
One of the simplest piecewise linear
functions is a contrast-stretching
transformation.
The idea behind contrast stretching is to
increase the dynamic range of the gray
levels in the image being processed.

15
Contrast stretching

16
Contrast stretching

17
 Gray-level slicing
Highlighting a specific range of gray
levels in an image often is desired.
There are several ways of doing level
slicing, but most of them are variations of
two basic themes.
One approach is to display a high value
for all gray levels in the range of interest
and a low value for all other gray levels.

18
 Gray-level slicing

0 255 0 255

19
 Bit-plane slicing
 Instead of highlighting gray-level ranges,
highlighting the contribution made to total
image appearance by specific bits might be
desired.
 Suppose that each pixel in an image is
represented b y 8 bits. Imagine that the image is
composed of eight 1-bit planes, ranging from
bit-plane 0, the least significant bit to bit-plane
7, the most significant bit.
20
 Bit-plane 0
(least significant)
1
1
0
0
1
1
10110011 0
1

Bit-plane 7
(most significant) 21
 Bit-plane slicing

Original Image

Bit-plane 7 Bit-plane 6 Bit-plane 4 Bit-plane 1

22
 Histogram Processing
 The histogram of a digital image with gray levels
in the range [0,L-1] is a discrete function
h(rk )  nk
Where rk is the kth gray level and
nk is the number of pixels in the image having
gray level rk

23
 HISTOGRAM
pixels
130

36 36
22

0 1 2 3 level
Image 16x14 = 224 pixels 24
 Role of Histogram Processing
The role of histogram processing in
image enhancement example
Dark image
Light image
Low contrast
High contrast

25
 Histogram Processing
Histogram Equalization
Histogram Matching(Specification)
Local Enhancement

26
 Initial part of discussion
Let r represent the gray levels of the
image to be enhanced.
Normally, we have used pixel values in
the interval [0,L-1], but now we assume
that r has been normalized to the interval
[0,1], with r=0 representing black and
r=1 representing white.

27
 Histogram Equalization
 We focus attention on transformations of the form
s  T ( r ) where 0  r  1
That produce a level s for every pixel value r in the
original image.

And we assume that the transformation function T(r)

satisfies the following conditions:
(a) T(r) is single-valued and monotonically increasing
in the interval 0 ≤ r ≤ 1; and
(b) 0 ≤ T(r) ≤ 1 for 0 ≤ r ≤ 1
28
 Reason of Condition
 The requirement in (a) that T9r) be single
valued is needed to guarantee that the inverse
transformation will exist, and the monotonicity
condition preserves the increasing order from
black to white in the output image.
 Condition (b) guarantees that the output gray
levels will be in the same range as the input
levels.

29
 Transformation function
Level s

sk=T(rk) T(r)

rk Level r
0 1

30
 Note
 The inverse transformation from s back to r is
denoted
s  T 1 (r ) where 0  r  1

There are some cases that even if T(r) satisfies conditions

(a) and (b), it is possible that the corresponding inverse
T-1(s) may fail to be single valued.

31
 Fundamental of random variable

PDF (probability density function) is the

probability of each element
CDF (cumulative distribution function) is
summation of the probability of the
element that value less than or equal this
element

32
 PDF
The PDF (probability density function) is
denoted by p(x)
ตัวอย่ าง การทอยลูกเต๋ า ความน่ าจะเป็ นที่ลกู เต๋ าจะขึน้ หน้ าต่ างๆ
เป็ นดังต่ อไปนี ้
pdf

1/6

1 2 3 4 5 6 หน้าลูกเต๋ า 33
 CDF
The CDF (cumulative density function) is
denoted by P(x)
ตัวอย่ าง การทอยลูกเต๋ า ความน่ าจะเป็ นที่ลกู เต๋ าจะขึน้ หน้ าต่ างๆ
เป็ นดังต่ อไปนี ้
cdf
1

1/6
1 2 3 4 5 6 หน้าลูกเต๋ า 34
 Relation between PDF and CDF
 PDF can find from this equation

d [ P( x)]
p( x)   P( x)
dx

 CDF can find from this equation

xn

P ( x  xn )   p( x)dx
x  x0

35
Idea of Histogram Equalization
 The gray levels in an image may be viewed as
random variables in the interval [0,1].
Let pr(r) denote the pdf of random variable r and
ps(s) denote the pdf of random variable s; if pr(r) and T(r)
are known and T-1(s) satisfies condition (a), the formula
should be
dr
p s ( s )  pr ( r )
ds
r
s  T (r )   pr ( w)dw
0
36
Idea of Histogram Equalization
 By Leibniz’s rule that the derivative of a definite
integral with respect to its upper limit is simply
the integrand evaluated at that limit.
ds dT (r )

dr dr
d  r

   pr ( w)dw
dr  0 
 pr ( r )

37
Idea of Histogram Equalization
 Substituting into the first equaltion

dr
p s ( s )  pr ( r )
ds
1
 pr ( r )
pr ( r )
 1, 0  s  1

38
Idea of Histogram Equalization
 The probability of occurrence of gray level rk in
an image is approximated by
nk
pk (rk )  k  0,1,2,..., L  1
n
k
sk  T (rk )   pr (rj )
j 0
k nj
 k  0,1,2,..., L  1
j 0 n
39
Histogram Equalization

40
Example for
Histogram Equalization
จงทำฮิ สโตแกรม อีควอไรเซชั่น ของ ค่ าสี จำนวนพิกเซล
(nj)
ภาพที่มีรายละเอียดต่ อไปนี ้
0 30
สูตร
k nj 1 50
sk   , k  0,1,.., L  1
j 0 n 2 100
L 1
3 1500
n   nj
j 0 4 2300
 30  50  100  1500  2300  4000  200  20 5 4000
 8200
6 200
7 20
41
Example for
Histogram Equalization
0 nj n0 30 0
s0      0.0037 
j 0 n n 8200 7
1 n
n  n 30  50 80 0
s1    0 1 
j
  0.0098 
j 0 n n 8200 8200 7
2 n
n n n 30  50  100 180 0
s2    0 1 2 
j
  0.022 
j 0 n n 8200 8200 7
3 n
30  50  100  1500 1680 1
s3   j    0.205 
j 0 n 8200 8200 7
4 n
30  50  100  1500  2300 3980 3
s4   
j
  0.485 
j 0 n 8200 8200 7
5 n
30  50  100  1500  2300  4000 7980 7
s5   
j
  0.973 
j 0 n 8200 8200 7
6 n
30  50  100  1500  2300  4000  200 8180 7
s6   j    0.998 
j 0 n 8200 8200 7
7 n
30  50  100  1500  2300  4000  200  20 8200 7
s7    42
j
 1
j 0 n 8200 8200 7
Example for
Histogram Equalization

43
Example for
Histogram Equalization(2)
ค่ าสี จำนวน
k k nj k nj
(k) พิกเซล(nj)
n
j 0
j n sk  ( L  1)
n
j 0 j 0

0 30 30 0.0037 0.0256
1 50 80 0.0098 0.0683
2 100 180 0.0220 0.1537
3 1500 1680 0.2050 1.4241
4 2300 3980 0.4854 3.3976
5 4000 7980 0.9732 6.8122
6 200 8180 0.9976 6.9830
7 20 8200 1.0000 7.0000
44
 Histogram Specification
can called “Histogram Matching”
Histogram equalization automatically
determines a transformation function that
seeks to produce an output image that
has a uniform histogram.
In particular, it is useful sometimes to be
able to specify the shape of the histogram
that we wish the processed image to have.

45
Meaning:
Histogram Specification
In this notation, r and z denote the gray
levels of the input and output (processed)
images, respectively.
We can estimate pr(r) from the given
image
While pz(z) is the specified probability
density function that we wish the output
image to have
46
Meaning:
Histogram Specification
Let s be a random variable with the
property s  T (r )  pr (w)dw  (1)
r

 0

Suppose that we define a random variable z

z
with property
G ( z )   p z (t )dt  s  (2)
0

Then two equations

z  G 1 ( s ) imply G(z)=T(r)
G 1[T (r )]  (3)

47
 Procedure:
Histogram Specification
1) Use Eq.(1) to obtain the transformation
function T(r)
2) Use Eq.(2) to obtain the transformation
function G(z)
3) Obtain the inverse transformation function G-
1

4) Obtain the output image by applying Eq.(3) to

all the pixels in the input image

48
Input Image

Specified histogram

49
Result form
Histogram Specification

50
 Local Enhancement
Normally, Transformation function based
on the content of an entire image.
Some cases it is necessary to enhance
details over small areas in an image.
The histogram processing techniques are
easily adaptable to local enhancement.

51
 Local Enhancement
Pixel-to-pixel translation Nonoverlapping region

52
Local Equalization

53
Enhancement Using
Arithmetic/Logic Operations
 Are performed on a pixel-by-pixel basis
between two or more images
 Logic operations are concerned with the ability
to implement the AND, OR, and NOT logic
operators because these three operators are
functionally complete.
 Arithmetic operations are concerned about
+,-,*, / and so on (arithmetic operators)

54
 Image Subtraction
The difference between two images f(x,y)
and h(x,y) expressed as
g(x,y) = f(x,y) – h(x,y)

55
56
 Image Averaging
Consider a noisy image g(x,y) formed by
the addition of noise to original image
f(x,y)
g ( x, y )  f ( x , y )   ( x , y ) where η(x,y) is noise

Where the assumption is that at every pair of

coordinates(x,y) the noise is uncorrelated and has zero
average value.
The objective of this procedure is to
reduce the noise content.
57
 Image Averaging
Let there are K different noisy images
If an image g ( x, y ) is formed by
averaging K different noisy images
K
1
g ( x, y ) 
K
 g ( x, y )
i 1
i

58
Image Averaging (Gray Scale)

1 image

2 5 10 20
images images images images

59
Image Averaging (Color Image)

(1) (2) (3)

Average image

60