1

of Contents
Over the next few lectures we will look at
image enhancement techniques working in
the spatial domain:
– What is image enhancement?
– Different kinds of image enhancement
– Histogram processing
– Point processing
– Neighbourhood operations
2
of A Note About Grey Levels
So far when we have spoken about image
grey level values we have said they are in
the range [0, 255]
– Where 0 is black and 255 is white
There is no reason why we have to use this
range
– The range [0,255] stems from display technologes
For many of the image processing
operations in this lecture grey levels are
assumed to be given in the range [0.0, 1.0]
3
of What Is Image Enhancement?
Image enhancement is the process of
making images more useful
The reasons for doing this include:
– Highlighting interesting detail in images
– Removing noise from images
– Making images more visually appealing
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
4
of
Image Enhancement Examples
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
5
of
Image Enhancement Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
6
of
Image Enhancement Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
7
of
Image Enhancement Examples (cont…)
8
of Spatial & Frequency Domains
There are two broad categories of image
enhancement techniques
– Spatial domain techniques
• Direct manipulation of image pixels
– Frequency domain techniques
• Manipulation of Fourier transform or wavelet
transform of an image
For the moment we will concentrate on
techniques that operate in the spatial
domain
9
of Image Histograms
The histogram of an image shows us the
distribution of grey levels in the image
Massively useful in image processing,
especially in segmentation
Frequencies

Grey Levels
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
10
Histogram Examples
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
11
Histogram Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
12
Histogram Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
13
Histogram Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
14
Histogram Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
15
Histogram Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
16
Histogram Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
17
Histogram Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
18
Histogram Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
19
Histogram Examples (cont…)
 20
of Histogram Examples (cont…)
A selection of images and
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

their histograms
Notice the relationships
between the images and
their histograms
Note that the high contrast
image has the most
evenly spaced histogram
21
of Contrast Stretching
We can fix images that have poor contrast
by applying a pretty simple contrast
specification
The interesting part is how do we decide on
this transformation function?
22
of Histogram Equalisation
Spreading out the frequencies in an image
(or equalising the image) is a simple way to
improve dark or washed out images
The formula for histogram
equalisation is given where sk = T (rk )
– rk: input intensity k

– sk: processed intensity =  pr ( r j )

j =1
– k: the intensity range
(e.g 0.0 – 1.0) k nj
– nj: the frequency of intensity j =
– n: the sum of all frequencies j =1 n
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
23
Equalisation Transformation Function
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
24

1
Equalisation Examples
 25
of Equalisation Transformation Functions

The functions used to equalise the images in

Images taken from Gonzalez & Woods, Digital Image Processing (2002)

the previous example

Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
26

2
Equalisation Examples
 27
of Equalisation Transformation Functions

The functions used to equalise the images in

Images taken from Gonzalez & Woods, Digital Image Processing (2002)

the previous example

Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
28

4
3
Equalisation Examples (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
29

4
3
Equalisation Examples (cont…)
 30
of Equalisation Transformation Functions

The functions used to equalise the images in

Images taken from Gonzalez & Woods, Digital Image Processing (2002)

the previous examples

31
of Summary
We have looked at:
– Different kinds of image enhancement
– Histograms
– Histogram equalisation
Now we will start to look at point processing
and some neighbourhood operations
32
of Contents
• we will look at image enhancement point
processing techniques:
– What is point processing?
– Negative images
– Thresholding
– Logarithmic transformation
– Power law transforms
– Grey level slicing
– Bit plane slicing
33
of
Basic Spatial Domain Image
Enhancement
•Most spatial domain enhancement
operations can be reduced to the form
Origin x
•g (x, y) = T[ f (x, y)]
•where f (x, y) is the
input image, g (x, y) is
the processed image (x, y)
and T is some
operator defined over
some neighbourhood
of (x, y) y Image f (x, y)
34
of Point Processing
•The simplest spatial domain operations
occur when the neighbourhood is simply the
pixel itself
•In this case T is referred to as a grey level
transformation function or a point processing
operation
•Point processing operations take the form
s=T(r)
•where s refers to the processed image pixel
value and r refers to the original image pixel
value
 35
of
Point Processing Example:
Negative Images
•Negative images are useful for enhancing
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

white or grey detail embedded in dark

regions of an image
– Note how much clearer the tissue is in the
negative image of the mammogram below

Original Negative
s = 1.0 - r
Image Image
36
of
Point Processing Example:
Negative Images (cont…)
Original Image Enhanced Image x
x

y Image f (x, y) y Image f (x, y)

s = intensitymax - r
 37
of
Point Processing Example:
Thresholding
•Thresholding transformations are
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

particularly useful for segmentation in which

we want to isolate an object of interest from
a background

1.0 r > threshold

s=
0.0 r <= threshold
38
of
Point Processing Example:
Thresholding (cont…)
Original Image Enhanced Image x
x

y Image f (x, y) y Image f (x, y)

1.0 r > threshold

s=
0.0 r <= threshold
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
39
Intensity Transformations
 40
of Basic Grey Level Transformations
•There are many different kinds of grey level
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

transformations
•Three of the most
common are shown
here
– Linear
• Negative/Identity
– Logarithmic
• Log/Inverse log
– Power law
• nth power/nth root
41
of Logarithmic Transformations
•The general form of the log transformation
is
s = c * log(1 + r)
•The log transformation maps a narrow
range of low input grey level values into a
wider range of output values
•The inverse log transformation performs the
opposite transformation
 42
of Logarithmic Transformations (cont…)

•Log functions are particularly useful when

Images taken from Gonzalez & Woods, Digital Image Processing (2002)

the input grey level values may have an

extremely large range of values
•In the following example the Fourier
transform of an image is put through a log
transform to reveal more detail

s = log(1 + r)
43
of Logarithmic Transformations (cont…)

Original Image Enhanced Image x

x

y Image f (x, y) y Image f (x, y)

s = log(1 + r)

We usually set c to 1
Grey levels must be in the range [0.0, 1.0]
 44
of Power Law Transformations
•Power law transformations have the following
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

form
s=c*rγ
•Map a narrow range
of dark input values
into a wider range of
output values or vice
versa
•Varying γ gives a whole
family of curves
45
of Power Law Transformations (cont…)
Original Image Enhanced Image x
x

y Image f (x, y) y Image f (x, y)

s=rγ

•We usually set c to 1

•Grey levels must be in the range [0.0, 1.0]
46
of Power Law Example
47
of Power Law Example (cont…)

γ = 0.6
1
Transformed Intensities

0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
Old Intensities
48
of Power Law Example (cont…)

γ = 0.4
1
0.9
Transformed Intensities

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
Original Intensities
49
of Power Law Example (cont…)

γ = 0.3
1
0.9
Transformed Intensities

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
Original Intensities
 50
of Power Law Example (cont…)
•The images to the
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

right show a
s = r 0.6
magnetic resonance
(MR) image of a
fractured human

s = r 0.4
spine
•Different curves
highlight different
detail
51
of Power Law Example
52
of Power Law Example (cont…)

γ = 5.0
1
0.9
Transformed Intensities

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
Original Intensities
 53
of Power Law Transformations (cont…)
•An aerial photo
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of a runway is
shown s = r 3.0
•This time
power law

s = r 4.0
transforms are
used to darken
the image
•Different curves
highlight
different detail
 54
of Gamma Correction
•Many of you might be familiar with gamma
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

correction of computer monitors

•Problem is that
display devices do
not respond linearly
to different
intensities
•Can be corrected
using a log
transform
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
55
More Contrast Issues
 56
of
Piecewise Linear Transformation
Functions
•Rather than using a well defined mathematical
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

function we can use arbitrary user-defined

transforms
•The images below show a contrast stretching
linear transform to add contrast to a poor
quality image
 57
of Gray Level Slicing
•Highlights a specific range of grey levels
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

– Similar to thresholding
– Other levels can be
suppressed or maintained
– Useful for highlighting features
in an image
58
of Image Enhancement
59
of Point Operations
60
of Non Linear Equations
• Square
• Square root
• Logarithm
• Power
61
of Contrast Stretching
62
of Range Normalization
63
of Clipping & Thresholding
64
of Multiple Thresholding
65
of Intensity Slicing
66
of
67
of Bitplane Slicing
68
of Histogram Stretching
69
of
70
of
71
of Histogram Stretching
72
of Histogram Sliding
73
of Histogram Equalization
74
of
75
of Histogram Specification
76
of
77
of
78
of
 79
of Bit Plane Slicing
•Often by isolating particular bits of the pixel
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

values in an image we can highlight

interesting aspects of that image
– Higher-order bits usually contain most of the
significant visual information
– Lower-order bits contain
subtle details
 80
of Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

[10000000] [01000000]

[00100000] [00001000]

[00000100] [00000001]
 81
of Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

[10000000] [01000000]

[00100000] [00001000]

[00000100] [00000001]
82
of Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
83
Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
84
Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
85
Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
86
Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
87
Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
88
Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
89
Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
90
Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

of
91
Bit Plane Slicing (cont…)
 92
of Bit Plane Slicing (cont…)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Reconstructed image
using only bit planes 8
and 7

Reconstructed image
using only bit planes 8, 7
and 6

Reconstructed image
using only bit planes 7, 6
and 5
93
of
94
of
95
of Summary
•We have looked at different kinds of point
processing image enhancement
•Next time we will start to look at
neighbourhood operations – in particular
filtering and convolution
96
of Histogram Processing
• Histogram of a digital image with gray levels in
the range [0,L-1] is a discrete function
h(rk) = nk
• Where
– rk : the kth gray level
– nk : the number of pixels in the image having gray
level rk
– h(rk) : histogram of a digital image with gray levels rk

96
97
of Normalized Histogram
• dividing each of histogram at gray level rk by the
total number of pixels in the image, n
p(rk) = nk / n
• For k = 0,1,…,L-1
• p(rk) gives an estimate of the probability of
occurrence of gray level rk
• The sum of all components of a normalized
histogram is equal to 1

97
98
of Histogram Processing
• Basic for numerous spatial domain
processing techniques
• Used effectively for image enhancement
• Information inherent in histograms also is
useful in image compression and
segmentation

98
99 h(rk) or p(rk)
of Example
rk

Dark image

Components of
histogram are
concentrated on the
low side of the gray
scale.
Bright image

Components of
histogram are
concentrated on the
high side of the gray
scale.
99
100
of Example

Low-contrast image

histogram is narrow
and centered toward
the middle of the
gray scale
High-contrast image

histogram covers broad

range of the gray scale
and the distribution of
pixels is not too far from
uniform, with very few
vertical lines being much
higher than
100 the others
101
of Histogram Equalization
• As the low-contrast image’s histogram is narrow
and centered toward the middle of the gray
scale, if we distribute the histogram to a wider
range the quality of the image will be improved.
• We can do it by adjusting the probability density
function of the original histogram of the image so
that the probability spread equally

101
102
of

Histogram transformation

s s = T(r)
• Where 0  r  1
• T(r) satisfies
– (a). T(r) is single-
sk= T(rk)
valued and
T(r)
monotonically
increasingly in the
interval 0  r  1
– (b). 0  T(r)  1 for
0r1
0 rk 1 r
102
103
of 2 Conditions of T(r)

• Single-valued (one-to-one relationship) guarantees

that the inverse transformation will exist
• Monotonicity condition preserves the increasing
order from black to white in the output image thus it
won’t cause a negative image
• 0  T(r)  1 for 0  r  1 guarantees that the output
gray levels will be in the same range as the input
levels.
• The inverse transformation from s back to r is
r = T -1(s) ;0s1
103
104
of Probability Density Function
• The gray levels in an image may be
viewed as random variables in the interval
[0,1]
• PDF is one of the fundamental descriptors
of a random variable

104
105
of Random Variables
• Random variables often are a source of
confusion when first encountered.
• This need not be so, as the concept of a
random variable is in principle quite
simple.

105
106
of Random Variables

• A random variable, x, is a real-valued function

defined on the events of the sample space, S.
• In words, for each event in S, there is a real
number that is the corresponding value of the
random variable.
• Viewed yet another way, a random variable maps
each event in S onto the real line.
• That is it. A simple, straightforward definition.
106
107
of Random Variables
• Part of the confusion often found in
connection with random variables is the
fact that they are functions.
• The notation also is partly responsible for
the problem.

107
108
of Random Variables
• In other words, although typically the
notation used to denote a random variable
is as we have shown it here, x, or some
other appropriate variable,
• to be strictly formal, a random variable
should be written as a function x(·) where
the argument is a specific event being
considered.

108
109
of Random Variables
• However, this is seldom done, and, in our
experience, trying to be formal by using
function notation complicates the issue
more than the clarity it introduces.
• Thus, we will opt for the less formal
notation, with the warning that it must be
keep clearly in mind that random variables
are functions.

109
110
of Random Variables
• Example:
– Consider the experiment of drawing a single
card from a standard deck of 52 cards.
– Suppose that we define the following events.
A: a heart; B: a spade; C: a club; and D: a
diamond, so that S = {A, B, C, D}.
– A random variable is easily defined by letting
x = 1 represent event A, x = 2 represent
event B, and so on.

110
111
of Random Variables
• As a second illustration,
– consider the experiment of throwing a single die and
observing the value of the up-face.
– We can define a random variable as the numerical
outcome of the experiment (i.e., 1 through 6), but
there are many other possibilities.
– For example, a binary random variable could be
defined simply by letting x = 0 represent the event
that the outcome of throw is an even number and
x = 1 otherwise.

111
112
of Random Variables
• Note
– the important fact in the examples just given
that the probability of the events have not
changed;
– all a random variable does is map events
onto the real line.

112
113
of Random Variables
• Thus far we have been concerned with
random variables whose values are
discrete.
• To handle continuous random variables
we need some additional tools.
• In the discrete case, the probabilities of
events are numbers between 0 and 1.

113
114
of Random Variables
• When dealing with continuous quantities
(which are not denumerable) we can no
longer talk about the "probability of an
event" because that probability is zero.
• This is not as unfamiliar as it may seem.

114
115
of Random Variables
• For example,
– given a continuous function we know that the
area of the function between two limits a and
b is the integral from a to b of the function.
– However, the area at a point is zero because
the integral from,say, a to a is zero.
– We are dealing with the same concept in the
case of continuous random variables.

115
116
of Random Variables
• Thus, instead of talking about the probability of a
specific value, we talk about the probability that
the value of the random variable lies in a
specified range.
• In particular, we are interested in the probability
that the random variable is less than or equal to
(or, similarly, greater than or equal to) a
specified constant a.
• We write this as
F(a) = P(x  a)
116
117
of Random Variables
• If this function is given for all values of a (i.e., −
 < a < ), then the values of random variable x
have been defined.
• Function F is called the cumulative probability
distribution function or simply the cumulative
distribution function (cdf).
• The shortened term distribution function also is
used.

117
118
of Random Variables

• Observe that the notation we have used makes

no distinction between a random variable and the
values it assumes.
• If confusion is likely to arise, we can use more
formal notation in which we let capital letters
denote the random variable and lowercase letters
denote its values.
• For example, the cdf using this notation is written
as
FX(x) = P(X  x)
118
119
of Random Variables
• When confusion is not likely, the cdf often
is written simply as F(x).
• This notation will be used in the following
discussion when speaking generally about
the cdf of a random variable.

119
120
of Random Variables

• Due to the fact that it is a probability,

the cdf has the following properties:
1. F(-) = 0
2. F() = 1
3. 0  F(x)  1
4. F(x1)  F(x2) if x1 < x2
5. P(x1 < x  x2) = F(x2) – F(x1)
6. F(x+) = F(x),
where x+ = x + , with  being a positive,
infinitesimally small number. 120
121
of Random Variables

The probability density function

(pdf or shortly called density function)
of random variable x is defined as the
derivative of the cdf:

dF ( x )
p( x ) =
dx
121
122
of Random Variables

The pdf satisfies the following properties:

122
123
of Random Variables
• The preceding concepts are applicable to
discrete random variables.
• In this case, there is a finite no. of events and we
talk about probabilities, rather than probability
density functions.
• Integrals are replaced by summations and,
sometimes, the random variables are
subscripted.
• For example, in the case of a discrete variable
with N possible values we would denote the
probabilities by P(xi), i=1, 2,…, N.

123
124
of Random Variables

• If a random variable x is transformed by a

monotonic transformation function T(x) to produce
a new random variable y,
• the probability density function of y can be obtained
from knowledge of T(x) and the probability density
function of x, as follows:
dx
p y ( y ) = px ( x )
dy
where the vertical bars signify the absolute value.
124
125
of Random Variables

• A function T(x) is monotonically increasing

if T(x1) < T(x2) for x1 < x2, and
• A function T(x) is monotonically decreasing
if T(x1) > T(x2) for x1 < x2.
• The preceding equation is valid if T(x) is an
increasing or decreasing monotonic
function.

125
126
of Applied to Image

• Let
– pr(r) denote the PDF of random variable r
– ps (s) denote the PDF of random variable s
• If pr(r) and T(r) are known and T-1(s) satisfies
condition (a) then ps(s) can be obtained using a
formula :
dr
p s(s) = p r (r)
ds 126
127
of Applied to Image

The PDF of the transformed variable s

is determined by
the gray-level PDF of the input image
and by
the chosen transformation function
127
128
of Transformation function

• A transformation function is a cumulative

distribution function (CDF) of random
variable r :
r
s = T ( r ) =  pr ( w )dw
0
where w is a dummy variable of integration

Note: T(r) depends on pr(r)

128
129
of
Cumulative
Distribution function
• CDF is an integral of a probability function
(always positive) is the area under the
function
• Thus, CDF is always single valued and
monotonically increasing
• Thus, CDF satisfies the condition (a)
• We can use CDF as a transformation
function

129
130
of Finding ps(s) from given T(r)

ds dT ( r )
=
dr dr
  dr
p s ( s ) = pr ( r )
r
d
=   pr ( w )dw  ds
dr  0 
1
= pr ( r ) = pr ( r )
pr ( r )
= 1 where 0  s  1
Substitute and yield
130
131
of ps(s)
• As ps(s) is a probability function, it must be
zero outside the interval [0,1] in this case
because its integral over all values of s
must equal 1.
• Called ps(s) as a uniform probability
density function
• ps(s) is always a uniform, independent of
the form of pr(r)

131
132
of r
s = T ( r ) =  pr ( w )dw
0

yields

Ps(s)

a random variable s 1
characterized by
a uniform probability
function 0
s
132
133
of
Discrete
transformation function

• The probability of occurrence of gray level

in an image is approximated by
nk
pr ( rk ) = where k = 0 , 1, ..., L- 1
n
• The discrete version of transformation
k
sk = T ( rk ) =  pr ( r j )
j =0
k nj
= where k = 0 , 1, ..., L- 1
j =0 n 133
134
of Histogram Equalization
• Thus, an output image is obtained by mapping
each pixel with level rk in the input image into a
corresponding pixel with level sk in the output
image
• In discrete space, it cannot be proved in general
that this discrete transformation will produce the
discrete equivalent of a uniform probability
density function, which would be a uniform
histogram

134
135
of Example

before after Histogram

equalization

135
136
of Example

before after Histogram

equalization

The quality is not

improved much
because the original
image already has a
broaden gray-level
scale

136
137
of Example
No. of pixels

6
2 3 3 2 5

4 2 4 3 4

3 2 3 5 3

2
2 4 2 4
1
Gray level
4x4 image
0 1 2 3 4 5 6 7 8 9
Gray scale = [0,9]
histogram
137
138
of

Gray
0 1 2 3 4 5 6 7 8 9
Level(j)
No. of
0 0 6 5 4 1 0 0 0 0
pixels
k

n
j =0
j 0 0 6 11 15 16 16 16 16 16

k nj 6 11 15 16 16 16 16 16
s= 0 0 / / / / / / / /
j =0 n
16 16 16 16 16 16 16 16
3.3 6.1 8.4
sx9 0 0 9 9 9 9 9
3 6 8
139
of Example
No. of pixels

6
3 6 6 3 5

8 3 8 6 4

6 3 6 9 3

2
3 8 3 8
1
Output image
0 1 2 3 4 5 6 7 8 9
Gray scale = [0,9] Gray level
Histogram equalization
139
140
of Note

• It is clearly seen that

– Histogram equalization distributes the gray level to
reach the maximum gray level (white) because the
cumulative distribution function equals 1 when
0  r  L-1
– If the cumulative numbers of gray levels are slightly
different, they will be mapped to little different or same
gray levels as we may have to approximate the
processed gray level of the output image to integer
number
– Thus the discrete transformation function can’t
guarantee the one to one mapping relationship
140
141
of Histogram Matching (Specification)

• Histogram equalization has a disadvantage

which is that it can generate only one type of
output image.
• With Histogram Specification, we can specify
the shape of the histogram that we wish the
output image to have.
• It doesn’t have to be a uniform histogram
141
142
Consider the continuous domain
of

Let pr(r) denote continuous probability density function of gray-

level of input image, r

Let pz(z) denote desired (specified) continuous probability

density function of gray-level of output image, z

Let s be a random variable with the property

r
s = T ( r ) =  pr ( w )dw Histogram equalization
0

Where w is a dummy variable of integration

142
143
Next, we define a random variable z with the property
of
z
g ( z ) =  pz ( t )dt = s Histogram equalization
0

Where t is a dummy variable of integration

thus

s = T(r) = G(z)

Therefore, z must satisfy the condition

z = G-1(s) = G-1[T(r)]

Assume G-1 exists and satisfies the condition (a) and (b)

We can map an input gray level r to output gray level z

143
144
of Procedure Conclusion
1. Obtain the transformation function T(r) by
calculating the histogram equalization of the
input image
r
s = T ( r ) =  pr ( w )dw
0

2. Obtain the transformation function G(z) by

calculating histogram equalization of the
desired density function
z
G ( z ) =  pz ( t )dt = s
0

144
145
of Procedure Conclusion
3. Obtain the inversed transformation
function G-1

z = G-1(s) = G-1[T(r)]

4. Obtain the output image by applying the

processed gray-level from the inversed
transformation function to all the pixels in
the input image

145
146
of Example

Assume an image has a gray level probability density function pr(r) as shown.

Pr(r)
 − 2r + 2 ;0  r  1
pr ( r ) = 
2
 0 ; elsewhere

1 r

 p ( w )dw = 1
0
r

0 1 2 r
146
147
of Example

We would like to apply the histogram specification with the desired

probability density function pz(z) as shown.

Pz(z)

 2z ;0  z  1
2 pz ( z ) = 
 0 ; elsewhere
1 z

z
 p ( w )dw = 1
0
z

0 1 2
147
148
of Step 1:

Obtain the transformation function T(r)

s=T(r) r
s = T ( r ) =  pr ( w )dw
0
1 r
=  ( −2 w + 2 )dw
One to one 0
mapping
r
function
= − w + 2w
2
0
r
0 1 = − r 2 + 2r
148
149
of Step 2:

Obtain the transformation function G(z)

z
G ( z ) =  ( 2w )dw
z
=z 2
=z 2
0
0

149
150
of Step 3:

Obtain the inversed transformation function G-1

G( z ) = T ( r )
z = − r + 2r
2 2

z = 2r − r 2

We can guarantee that 0  z 1 when 0  r 1

150
151
of Discrete formulation

k
sk = T ( rk ) =  pr ( r j )
j =0
k nj
= k = 0 ,1,2 ,..., L − 1
j =0 n
k
G ( z k ) =  pz ( z i ) = sk k = 0 ,1,2 ,..., L − 1
i =0

z k = G −1 T ( rk )
=G −1
sk  k = 0 ,1,2 ,..., L 151
−1
152
of Example

Image is dominated by large, dark areas,

resulting in a histogram characterized by
a large concentration of pixels in pixels in
the dark end of the gray scale
Image of Mars moon 152
153
of Image Equalization

Result image
after histogram
equalization
Transformation function
Histogram of the result image
for histogram equalization
The histogram equalization doesn’t make the result image look better than
the original image. Consider the histogram of the result image, the net
effect of this method is to map a very narrow interval of dark pixels into
the upper end of the gray scale of the output image. As a consequence, the
output image is light and has a washed-out appearance. 153
154
of Solve the problem
Histogram Equalization

Since the problem with the

transformation function of the
histogram equalization was
caused by a large concentration
of pixels in the original image with
Histogram Specification
levels near 0

a reasonable approach is to
modify the histogram of that
image so that it does not have
this property
154
155
of

Histogram Specification

• (1) the transformation

function G(z) obtained
from
k
G ( z k ) =  pz ( z i ) = sk
i =0

k = 0 ,1,2 ,..., L − 1

• (2) the inverse

transformation G-1(s)

155
156
of

Result image and its histogram

The output image’s histogram

Notice that the output

histogram’s low end has
After applied the
Original image histogram shifted right toward the
equalization lighter region of the gray
scale as desired.
156
157
of Note
• Histogram specification is a trial-and-error
process
• There are no rules for specifying
histograms, and one must resort to
analysis on a case-by-case basis for any
given enhancement task.

157
158
of Note
• Histogram processing methods are global
processing, in the sense that pixels are
modified by a transformation function
based on the gray-level content of an
entire image.
• Sometimes, we may need to enhance
details over small areas in an image,
which is called a local enhancement.

158
159 a) Original image
of (slightly blurred to
reduce noise)
b) global histogram
equalization (enhance
noise & slightly
increase contrast but
the construction is
not changed)
c) local histogram
equalization using
7x7 neighborhood
(reveals the small
squares inside larger
ones of the original
image.
(a) (b) (c)

• define a square or rectangular neighborhood and move the center of

this area from pixel to pixel.
• at each location, the histogram of the points in the neighborhood is
computed and either histogram equalization or histogram specification
transformation function is obtained.
• another approach used to reduce computation is to utilize
nonoverlapping regions, but it usually produces an undesirable
checkerboard effect. 159
160
of Explain the result in c)

• Basically, the original image consists of many

small squares inside the larger dark ones.
• However, the small squares were too close in
gray level to the larger ones, and their sizes were
too small to influence global histogram
equalization significantly.
• So, when we use the local enhancement
technique, it reveals the small areas.
• Note also the finer noise texture is resulted by the
local processing using relatively small
neighborhoods.
160
161
of
Enhancement using Arithmetic/Logic
Operations
• Arithmetic/Logic operations perform on
pixel by pixel basis between two or more
images
• except NOT operation which perform only
on a single image

161
162
of Logic Operations
• Logic operation performs on gray-level
images, the pixel values are processed as
binary numbers
• light represents a binary 1, and dark
represents a binary 0
• NOT operation = negative transformation

162
163
of

Example of AND Operation

original image AND image result of AND

mask operation

163
164
of

Example of OR Operation

original image OR image result of OR

mask operation

164
165
of Image Subtraction

g(x,y) = f(x,y) – h(x,y)

• enhancement of the differences between

images

165
166 a b
of
c d
Image Subtraction
• a). original fractal image
• b). result of setting the four lower-
order bit planes to zero
– refer to the bit-plane slicing
– the higher planes contribute
significant detail
– the lower planes contribute more
to fine detail
– image b). is nearly identical
visually to image a), with a very
slightly drop in overall contrast
due to less variability of the gray-
level values in the image.
• c). difference between a). and b).
(nearly black)
• d). histogram equalization of c).
(perform contrast stretching
transformation)

166
167
of

Mask mode radiography

• h(x,y) is the mask, an X-ray
image of a region of a patient’s
body captured by an intensified
TV camera (instead of
traditional X-ray film) located
opposite an X-ray source
• f(x,y) is an X-ray image taken
after injection a contrast
medium into the patient’s
mask image an image (taken after bloodstream
injection of a contrast • images are captured at TV
medium (iodine) into the rates, so the doctor can see
bloodstream) with mask how the medium propagates
Note: subtracted out. through the various arteries in
• the background is dark because it the area being observed (the
doesn’t change much in both images. effect of subtraction) in a movie
• the difference area is bright because it showing mode.
has a big change 167
168
of Note

• We may have to adjust the gray-scale of the subtracted image

to be [0, 255] (if 8-bit is used)
– first, find the minimum gray value of the subtracted image
– second, find the maximum gray value of the subtracted
image
– set the minimum value to be zero and the maximum to be
255
– while the rest are adjusted according to the interval
[0, 255], by timing each value with 255/max
• Subtraction is also used in segmentation of moving pictures to
track the changes
– after subtract the sequenced images, what is left should be
the moving elements in the image, plus noise
168
169
of Image Averaging
• consider a noisy image g(x,y) formed by
the addition of noise (x,y) to an original
image f(x,y)

g(x,y) = f(x,y) + (x,y)

169
170
of Image Averaging
• if noise has zero mean and be
uncorrelated then it can be shown that if

g ( x, y) = image formed by averaging

K different noisy images

K
1
g ( x, y ) =
K
 g ( x, y )
i =1
i

170
171
of Image Averaging
• then

1 2
 2
g ( x, y ) =   ( x, y )
K
 2
g ( x, y ) , 2
 ( x, y ) = variances of g and 

if K increase, it indicates that the variability (noise) of the pixel at each location
(x,y) decreases.

171
172
of Image Averaging
• thus

E{g ( x, y)} = f ( x, y)

E{g ( x, y)} = expected value of g

(output after averaging)

= original image f(x,y)

172
173
of Image Averaging
• Note: the images gi(x,y) (noisy images)
must be registered (aligned) in order to
avoid the introduction of blurring and other
artifacts in the output image.

173
174 a b
of
c d
Example
e f

• a) original image
• b) image corrupted by
additive Gaussian noise
with zero mean and a
standard deviation of 64
gray levels.
• c). -f). results of averaging
K = 8, 16, 64 and 128
noisy images
174
175
of Spatial Filtering

• use filter (can also be called as

mask/kernel/template or window)
• the values in a filter subimage are referred
to as coefficients, rather than pixel.
• our focus will be on masks of odd sizes,
e.g. 3x3, 5x5,…

175
176
of Spatial Filtering Process

• simply move the filter mask from point to

point in an image.
• at each point (x,y), the response of the filter
at that point is calculated using a
predefined relationship.
R = w1 z1 + w2 z2 + ... + wmn zmn
mn
=  wi zi
i =i 176
177
of Linear Filtering

• Linear Filtering of an image f of size MxN

filter mask of size mxn is given by the
expression
a b
g ( x, y ) =   w(s, t ) f ( x + s, y + t )
t =− a t =−b

where a = (m-1)/2 and b = (n-1)/2

To generate a complete filtered image this equation must be applied for x =

0, 1, 2, … , M-1 and y = 0, 1, 2, … , N-1
177
178
of Smoothing Spatial Filters

• used for blurring and for noise reduction

• blurring is used in preprocessing steps,
such as
– removal of small details from an image prior to
object extraction
– bridging of small gaps in lines or curves
• noise reduction can be accomplished by
blurring with a linear filter and also by a
nonlinear filter
178
179
of Smoothing Linear Filters

• output is simply the average of the pixels

contained in the neighborhood of the filter
mask.
• called averaging filters or lowpass filters.

179
180
of Smoothing Linear Filters
• replacing the value of every pixel in an image by
the average of the gray levels in the
neighborhood will reduce the “sharp” transitions
in gray levels.
• sharp transitions
– random noise in the image
– edges of objects in the image
• thus, smoothing can reduce noises (desirable)
and blur edges (undesirable)

180
181
of 3x3 Smoothing Linear Filters

box filter weighted average

the center is the most important and other

pixels are inversely weighted as a function of
their distance from the center of the mask
181
182
of Weighted average filter
• the basic strategy behind weighting the
center point the highest and then reducing
the value of the coefficients as a function
of increasing distance from the origin is
simply an attempt to reduce blurring in
the smoothing process.

182
183
of General form : smoothing mask

• filter of size mxn (m and n odd)

a b

  w(s, t ) f ( x + s, y + t )
g ( x, y ) = s = − at = − b
a b

  w(s, t )
s = − at = − b

summation of all coefficient of the mask

183
184 a b
of
c d
Example
e f

• a). original image 500x500 pixel

• b). - f). results of smoothing with
square averaging filter masks of
size n = 3, 5, 9, 15 and 35,
respectively.
• Note:
– big mask is used to eliminate small
objects from an image.
– the size of the mask establishes the
relative size of the objects that will
be blended with the background.

184
185
of Example

original image result after smoothing result of thresholding

with 15x15 averaging mask
we can see that the result after smoothing and thresholding, the remains are the
largest and brightest objects in the image.
185
186
of
Order-Statistics Filters (Nonlinear
Filters)
• the response is based on ordering
(ranking) the pixels contained in the image
area encompassed by the filter
• example
– median filter : R = median{zk |k = 1,2,…,n x n}
– max filter : R = max{zk |k = 1,2,…,n x n}
– min filter : R = min{zk |k = 1,2,…,n x n}
• note: n x nis the size of the mask

186
187
of Median Filters
• replaces the value of a pixel by the median of
the gray levels in the neighborhood of that pixel
(the original value of the pixel is included in the
computation of the median)
• quite popular because for certain types of
random noise (impulse noise  salt and pepper
noise) , they provide excellent noise-reduction
capabilities, with considering less blurring than
linear smoothing filters of similar size.

187
188
of Median Filters

• forces the points with distinct gray levels to be

more like their neighbors.
• isolated clusters of pixels that are light or dark
with respect to their neighbors, and whose area is
less than n2/2 (one-half the filter area), are
eliminated by an n x n median filter.
• eliminated = forced to have the value equal the
median intensity of the neighbors.
• larger clusters are affected considerably less

188
189
of Example : Median Filters

189
190
of Sharpening Spatial Filters
• to highlight fine detail in an image
• or to enhance detail that has been blurred,
either in error or as a natural effect of a
particular method of image acquisition.

190
191
of Blurring vs. Sharpening

• as we know that blurring can be done in

spatial domain by pixel averaging in a
neighbors
• since averaging is analogous to integration
• thus, we can guess that the sharpening must
be accomplished by spatial differentiation.

191
192
of Derivative operator

• the strength of the response of a derivative

operator is proportional to the degree of
discontinuity of the image at the point at which the
operator is applied.
• thus, image differentiation
– enhances edges and other discontinuities (noise)
– deemphasizes area with slowly varying gray-level
values.

192
193
of First-order derivative
• a basic definition of the first-order
derivative of a one-dimensional function
f(x) is the difference

f
= f ( x + 1) − f ( x)
x

193
194
of Second-order derivative
• similarly, we define the second-order
derivative of a one-dimensional function
f(x) is the difference

 f
2
= f ( x + 1) + f ( x − 1) − 2 f ( x)
x 2

194
195
of
First and Second-order derivative of
f(x,y)
• when we consider an image function of
two variables, f(x,y), at which time we will
dealing with partial derivatives along the
two spatial axes.

f ( x, y ) f ( x, y ) f ( x, y )
Gradient operator f = = +
xy x y
Laplacian operator
 f ( x, y)  f ( x, y)
2 2
(linear operator)  f =
2
+
x 2
y
195
2
196
of Discrete Form of Laplacian

from  f 2
= f ( x + 1, y ) + f ( x − 1, y ) − 2 f ( x, y )
x 2

 f
2
= f ( x, y + 1) + f ( x, y − 1) − 2 f ( x, y)
y 2

yield,

 f = [ f ( x + 1, y ) + f ( x − 1, y )
2

+ f ( x, y + 1) + f ( x, y − 1) − 4 f ( x, y )]
196
197
of Result Laplacian mask

197
198
of
Laplacian mask implemented an
extension of diagonal neighbors

198
199
of
Other implementation of Laplacian
masks

give the same result, but we have to keep in mind that when combining
(add / subtract) a Laplacian-filtered image with another image.

199
200
of Effect of Laplacian Operator
• as it is a derivative operator,
– it highlights gray-level discontinuities in an
image
– it deemphasizes regions with slowly varying
gray levels
• tends to produce images that have
– grayish edge lines and other discontinuities,
all superimposed on a dark,
– featureless background.

200
201
of
Correct the effect of featureless
background

• easily by adding the original and Laplacian

image.
• be careful with the Laplacian filter used
if the center coefficient
of the Laplacian mask is
 f ( x, y ) −  2 f ( x , y ) negative
g ( x, y ) = 
 f ( x , y ) +  2
f ( x, y )
if the center coefficient
of the Laplacian mask is
positive
201
202
of

Example
• a). image of the North
pole of the moon
• b). Laplacian-filtered
image with
1 1 1
1 -8 1
1 1 1

• c). Laplacian image

scaled for display
purposes
• d). image enhanced by
addition with original
image 202
203
of Mask of Laplacian + addition
• to simply the computation, we can create a
mask which do both operations, Laplacian
Filter and Addition the original image.

203
204
of Mask of Laplacian + addition

g ( x, y ) = f ( x, y ) − [ f ( x + 1, y ) + f ( x − 1, y )
+ f ( x, y + 1) + f ( x, y − 1) + 4 f ( x, y )]
= 5 f ( x, y ) − [ f ( x + 1, y ) + f ( x − 1, y )
+ f ( x, y + 1) + f ( x, y − 1)]

0 -1 0
-1 5 -1
0 -1 0
204
205
of Example

205
206
of Note
 f ( x, y ) −  2 f ( x , y )
g ( x, y ) = 
 f ( x , y ) +  2
f ( x, y )

0 -1 0 0 0 0 0 -1 0
-1 5 -1 = 0 1 0 + -1 4 -1
0 -1 0 0 0 0 0 -1 0

0 -1 0 0 0 0 0 -1 0
-1 9 -1 = 0 1 0 + -1 8 -1
0 -1 0 0 0 0 0 -1 0

206
207
of Unsharp masking

f s ( x, y) = f ( x, y) − f ( x, y)
sharpened image = original image – blurred image

• to subtract a blurred version of an image

produces sharpening output image.

207
208
of High-boost filtering

f hb ( x, y) = Af ( x, y) − f ( x, y)

f hb ( x, y ) = ( A − 1) f ( x, y ) − f ( x, y ) f ( x, y )
= ( A − 1) f ( x, y ) − f s ( x, y )
• generalized form of Unsharp masking
• A1
208
209
of High-boost filtering

f hb ( x, y) = ( A −1) f ( x, y) − f s ( x, y)
• if we use Laplacian filter to create sharpen
image fs(x,y) with addition of original image

 f ( x, y ) −  f ( x, y )
2
f s ( x, y ) = 
 f ( x, y ) +  f ( x, y )
2

209
210
of High-boost filtering

if the center coefficient

• yields of the Laplacian mask is
negative

 Af ( x, y ) −  f ( x, y )
2
f hb ( x, y ) = 
 Af ( x, y ) +  f ( x, y )
2

if the center coefficient

of the Laplacian mask is
positive
210
211
of High-boost Masks

◼ A1
◼ if A = 1, it becomes “standard” Laplacian
sharpening 211
212
of Example

212
213  f 
of Gradient Operator
Gx   x 
f =   =  f 
G y   
 y 

• first derivatives are implemented using the

magnitude of the gradient.
1
f = mag (f ) = [G + G ] 2
x
2
y
2
commonly approx.
1
 f  2  f  2  2

=   +   
 x   y  
f  Gx + G y
the magnitude becomes nonlinear
213
214 z2 z3
z1
of Gradient Mask
z4 z5 z6
z7 z8 z9

• simplest approximation, 2x2

Gx = ( z8 − z5 ) and G y = ( z 6 − z5 )
1 1
f = [G + G ]
2
x
2
y
2
= [( z8 − z5 ) + ( z6 − z5 ) ]
2 2 2

f  z8 − z 5 + z 6 − z 5

214
215 z2 z3
z1
of Gradient Mask
z4 z5 z6
z7 z8 z9

• Roberts cross-gradient operators, 2x2

Gx = ( z9 − z5 ) and Gy = ( z8 − z6 )
1 1
f = [G + G ]
2
x
2
y
2
= [( z9 − z5 ) + ( z8 − z6 ) ]
2 2 2

f  z 9 − z 5 + z8 − z 6

215
216 z2 z3
z1
of Gradient Mask
z4 z5 z6
z7 z8 z9

• Sobel operators, 3x3

Gx = ( z7 + 2 z8 + z9 ) − ( z1 + 2 z 2 + z3 )
G y = ( z3 + 2 z6 + z9 ) − ( z1 + 2 z 4 + z7 )
f  Gx + G y
the weight value 2 is to achieve
smoothing by giving more
important to the center point

216
217
of Note
• the summation of coefficients in all masks
equals 0, indicating that they would give a
response of 0 in an area of constant gray
level.

217
218
of Example

218
219
of
Example of Combining Spatial
Enhancement Methods
• want to sharpen the
original image and
bring out more skeletal
detail.
• problems: narrow
dynamic range of gray
level and high noise
content makes the
image difficult to
enhance
219
220
of
Example of Combining Spatial
Enhancement Methods
• solve :
1. Laplacian to highlight fine detail
2. gradient to enhance prominent
edges
3. gray-level transformation to
increase the dynamic range of
gray levels

220
221
of

221
222
of

222