Chapter Three

Spatial Domain Image Processing

1
 3.1. Spatial Processing of Digital
Images
 The term spatial domain refers to the image
plane itself, and image processing methods in
this category are based on direct
manipulation of pixels in an image.
 Two principal categories of spatial processing
are intensity transformations and spatial
filtering.
 Intensity transformations operate on single
pixels of an image for tasks such as contrast
manipulation and image thresholding.
2
…cont’d

 Spatial filtering performs operations on the

neighborhood of every pixel in an image.
Examples of spatial filtering include image
smoothing and sharpening.

3
The Basics of Intensity Transformations
and Spatial Filtering

 The spatial domain processes are

mathematically expressed as:

where f(x,y) is an input image, g(x,y) is the

output image, and T is an operator on f
defined over a neighborhood of point (x,y).
 The operator can be applied to the pixels of a
single image or to the pixels of a set of
images

4
…cont’d

 Figure shows the basic implementation of the

above equation on a single image.

5
…cont’d

 The process that the above figure illustrates

consists of moving the center of the
neighborhood from pixel to pixel, and
applying the operator T to the pixels in the
neighborhood to yield an output value at that
location.
 Thus, for any specific location (x0,y0), the
value of the output image g at those
coordinates is equal to the result of applying
T to the neighborhood with origin at (x0,y0)in f.
6
…cont’d

 The smallest possible neighborhood is of size

1x1.
 In this case, g depends only on the value of f
at a single point (x,y) and T becomes an
intensity (also called a gray-level, or
mapping) transformation function of the form
s = T(r)
where, s and r to denote the intensity of g
and f respectively at any point (x,y).

7
…cont’d

 For example, contrast stretching and

thresholding function.

Contrast stretching Thresholding

8
…cont’d

 Contrast stretching is a linear mapping

function used to manipulate the contrast of an
image.
 Thresholding produces a binary image

9
3.2. Basic Intensity Transformation
Functions
 There are three basic types of intensity
transformation functions:
 Linear (negative and identity transformations),
 Logarithmic (log and inverse-log
transformations), and
 Power-law (nth power and nth root
transformations).
 The identity function is the trivial case in which
the input and output intensities are identical.

10
…cont’d

11
Image Negatives

 The negative of an image with intensity levels

in the range [ 0, L-1] is obtained by using the
negative transformation function which has
the form: s = L – 1 – r
 This type of processing is used, for example,
in enhancing white or gray detail embedded
in dark regions of an image, especially when
the black areas are dominant in size.

12
…cont’d

Image Negative Image

13
Log Transformations

 The general form of the log transformation in

is s = c log(1 + r)
where c is a constant and it is assumed that r ≥ 0.
 The logarithmic (log) transformation is used to
compress the dynamic range of an image.
 Log transformation maps a narrow range of
low intensity values in the input into a wider
range of output levels and the opposite is true
for higher values of input grayscale levels.

14
…cont’d

 In other words, values of dark pixels in an

image are expanded, while the high
grayscale values are compressed.
 The inverse log transformation does the
opposite.
 As shown in Figure, log transformation is
used to view Fourier transformed images, as
their dynamic ranges are very high.
 Log transformations show the details that are
not visible due to a large dynamic range of
pixel values. 15
…cont’d

Fourier spectrum displayed as a Result of applying the log

grayscale image. transformation with c = 1

16
 Power-law (Gamma)
Transformations
 Power-law transformations have the form

where c and are positive constants.

 As with log transformations, power-law curves
with fractional values of map a narrow
range of dark input values into a wider range
of output values, with the opposite being true
for higher values of input levels.

17
…cont’d

 Power-law transformations are useful for

general-purpose contrast manipulation.

18
 Piecewise Linear Transformation
Functions
 Complementary approach to Linear,
Logarithmic, and Power-law transformations.
 The advantage of these functions is that the
form of piecewise functions can be arbitrarily
complex.
 The main disadvantage of these functions is
that their specification requires considerable
user input.
 Example: Contrast Stretching, Intensity-Level
Slicing, Bit-Plane Slicing
19
Contrast Stretching
 Low-contrast images can result from poor
illumination, lack of dynamic range in the
imaging sensor, or even the wrong setting of a
lens aperture during image acquisition.
 Contrast stretching expands the range of
intensity levels in an image so that it spans the
ideal full intensity range of the recording
medium or display device.

20
…cont’d

Piecewise linear A low contrast Result of

transformation electron image contrast stretching.
function.

Contrast stretching is obtained by setting (r1 , s1) = (rmin, 0) and (r2

, s2 ) = (rmax, L-1), where rmin and rmax denote the minimum and
maximum intensity levels in the input image 21
Intensity-Level Slicing
 Highlight a specific range of intensities in an
image.
 Intensity-level slicing, can be implemented in
several ways, but most are variations of two
basic themes.
 One approach is to display in one value (say,
white) all the values in the range of interest
and in another (say, black) all other intensities.
 This transformation, produces a binary image.
22
…cont’d

 The second approach, brightens (or darkens)

the desired range of intensities, but leaves all
other intensity levels in the image unchanged.

Original First approach Second approach

23
Bit-Plane Slicing
 Pixel values are integers composed of bits.
For example, values in a 256-level grayscale
image are composed of 8 bits (one byte).
 Instead of highlighting intensity-level ranges,
Bit-Plane Slicing highlight the contribution
made to total image appearance by specific
bits.

24
Bit-planes of an 8-bit image.
…cont’d

25
3.3 Histogram Processing

Image Histogram
 An image histogram is a type of histogram that
acts as a graphical representation of the tonal
distribution in a digital image.
 It plots the number of pixels for each tonal
value
 The horizontal axis of the graph represents the
tonal variations, while the vertical axis
represents the total number of pixels in that
particular tone.
26
…cont’d

Image

27
Image Histogram
…cont’d

 Let rk, for k = 0 1 2 ,…, L - 1 denote the

intensities of an L-level digital image, f(x, y).
 The unnormalized histogram of f is defined as

where nk is the number of pixels in f with

intensity rk, and the subdivisions of the
intensity scale are called histogram bins.
 The normalized histogram of f is defined as

Where, M = image rows and N = image columns 28

…cont’d

 Normalized histogram referred to as

histograms or image histograms.
 The sum of p(rk) for all values of k is always 1.
 Histogram shape is related to image
appearance.

29
Histogram Equalization

 Histogram equalization is used for equalizing

all the pixel values of an image.
 Transformation is done in such a way that
uniform flattened histogram is produced.
 Histogram equalization increases the
dynamic range of pixel values and makes an
equal count of pixels at each level which
produces a flat histogram with high contrast
image.

30
…cont’d

Before equalization After equalization

31
 For continuous intensity value, a histogram
equalization or histogram linearization
transformation is given by:

where L is the number of possible intensity levels

in the image and the integral on the right side is
the cumulative distribution function (CDF) of
random variable r.
 Recall that the probability of occurrence of
intensity level rk in a digital image is
approximated by:
32
…cont’d

• For discrete intensity value, a histogram

equalization or histogram linearization
transformation is given by:

33
3.4 Spatial Filtering

 Filter is the process that moves some

unwanted components or small details in an
image.
 Spatial filtering modifies an image by
replacing the value of each pixel by a function
of the values of the pixel and its neighbors.
 There are two types of spatial filtering:
 Linear spatial filter – the operation performed
on the image pixels is linear
 Nonlinear spatial filter the operation performed
on the image pixels is nonlinear 34
Linear spatial filtering

 A linear spatial filter performs a sum-of-

products operation between an image f and a
filter kernel (mask), w.
 The kernel is an array whose size defines the
neighborhood of operation, and whose
coefficients determine the nature of the filter.
 The filter mask may be 3x3, 5x5, or 7x7 mask
 At any point (x, y) in the image, the response,
g(x, y), of the filter is the sum of products of
the kernel coefficients and the image pixels
encompassed by the kernel: 35
…cont’d

 In general, linear spatial filtering of an image

of size M x N with a kernel of size m x n is
given by the expression:

 For a kernel of size m x n , we assume that

m=2a +1 and n=2b+1, where a and b are
nonnegative integers. 36
…cont’d

37
 Spatial Correlation and
Convolution
 Correlation consists of moving the center of a
kernel over an image, and computing the sum
of products at each location.
 Spatial convolution are the same, except that
the correlation kernel is rotated by 180°.
 The correlation of a kernel w of size m x n
with an image f(x,y), is given by

38
…cont’d

 In a similar manner, the convolution of a

kernel w of size m n × with an image f(x,y), is
defined as

39
Types of Spatial Filter

 Spatial filters can be classified by effect into:

 Smoothing spatial filters – also called lowpass
filters
 Sharpening spatial filters – also called
highpass filters

40
Smoothing spatial filters

 Smoothing filters are used for blurring and for

nose reduction.
 Blurring used as preprocessing such as
removal of small details from image.
 Noise reduction is blurring with linear or
nonlinear filter.
 Smoothing filters include:
 Averaging
 Order statistics(non linear)
41
Lowpass Box Filter Kernels

 The simplest, lowpass filter kernel is the box

kernel, whose coefficients have the same
value (typically 1).
 An mxn box filter is an mxn array of 1’s, with
a normalizing constant in front, whose value
is 1 divided by the sum of the values of the
coefficients (i.e., 1/mn when all the
coefficients are 1’s).

3x3 box kernel 42

 …cont’d

Origil Image Filtered image 3x3

Filtered Image Filtered image

11x11 21x21

Image averaging 43
Lowpass Gaussian Filter Kernels

 Gaussian kernels is obtained by:

 A 3x3 guassian kernel for K = 1 and

44
…cont’d

(a) A test pattern of size 1024x1024. (b) Result of

lowpass filtering the pattern with a Gaussian kernel of
size 21x21 , with (c) Result of using a kernel of
size 43x43, with

45
Order-statistic (Nonlinear) Filters

 Order-statistic filters are nonlinear spatial

filters whose response is based on ordering
(ranking) the pixels contained in the region
encompassed by the filter.
 Smoothing is achieved by replacing the value
of the center pixel with the value determined
by the ranking result.
 It includes:
 Median filters
 Min filters
 Max filters 46
…cont’d
Median filter
 Replaces the value of the center pixel by the
median of the intensity values in the
neighborhood of that pixel
 It is best for noise reduction
Max filter
 Useful for finding the brightest points in an
image or for eroding dark areas adjacent to
light regions.
 Replaces the value of the center pixel by the
maximum of the intensity values in the
47
…cont’d

Min filter
 Replaces the value of the center pixel by the
minimum of the of the intensity values in the
neighborhood of that pixel
 Useful for finding the darkest points in an
image

48
Sharpening Spatial Filters

 Sharpening highlights transitions in intensity.

 Applicable in:
 Electronic printing,
 Medical imaging,
 Industrial inspection, and
 Autonomous guidance in military systems.

49
Foundations

 Sharpening filters are based on first- and

second-order derivatives
First Derivative:
 Must be zero in areas of constant intensity.
 Must be nonzero at the onset of an intensity
step or ramp.
 Must be nonzero along intensity ramps.
 A basic definition of the first-order derivative of
a one-dimensional function f(x) is the
difference
50
…cont’d

Second Derivatives:
 Must be zero in areas of constant intensity.
 Must be nonzero at the onset and end of an
intensity step or ramp.
 Must be zero along intensity ramps.
 We define the second-order derivative of f(x )
as the difference

51
…cont’d

52
Image Sharpening – the Laplacian

 This approach uses the second order

derivatives for constructing a filter mask
(kernel)
 The Laplacian for the image function f(x,y) is

 In the x-direction, we have

 In the y- direction, we have

53
…cont’d

 It follows from the preceding three equations

that the discrete Laplacian of two variables is

(a) (b) (c) (d)

Laplacian Kernels
54
…cont’d

 If a negative center coefficient is used , then

we subtract the Laplacian image from the
original to obtain a sharpened result.
 Thus, the basic way in which we use the
Laplacian for image sharpening is

where f(x,y) and g(x,y) are the input and

sharpened images, respectively. We let c = −1 if
the Laplacian kernels in above figure (a) or (b) is
used, & c = 1 if either of the other two kernels is
used. 55
 …cont’d

Original image Laplacian image

using kernel (a)

Image sharpend Image with

using c = -1 kernel (B)

56
Image Sharpening – the Gradient

 First derivatives in image processing are

implemented using the magnitude of the
gradient.
 The gradient of an image f at coordinates
(x,y) is defined as the two dimensional
column vector

57
…cont’d

 The magnitude (length) of vector , denoted

as M(x,y) (the vector norm notation is also
used frequently), where

is the value at (x,y) of the rate of change in the

direction of the gradient vector.
 M(x,y) is an image of the same size as the
original, created when x and y are allowed to
vary over all pixel locations in f. It is called
gradient image
58
 Chapter Four

Frequency Domain Image Processing

1
 4.1 Introduction to Frequency
Domain
 In spatial domain, we deal with images as it
is. The value of the pixels of the image
change with respect to scene.
 Whereas in frequency domain, we deal with
the rate at which the pixel values are
changing in spatial domain.
 We use the Fourier transform to go from the
spatial to the frequency domain; to return to
the spatial domain we use the inverse Fourier
transform.
2
4.2 Fourier series

 A function f(t) of a continuous variable, t, that

is periodic with a period, T, can be expressed
as the sum of sines and cosines multiplied by
appropriate coefficients. This sum, known as
a Fourier series, has the form

where

are the coefficients.

3
For example

 The function at the bottom is the sum of the four

functions above it

4
Fourier Transform

 Functions that are not periodic can be

expressed as the integral of sines and/or
cosines multiplied by a weighting function.
 This sum, known as a Fourier transform, has
the form

 Conversely, given , we can obtain f(t)

back using the inverse Fourier transform,
written as
5
…cont’d

 Fourier transformation is used in a wide

range of application such as image filtering,
image compression, Image analysis and
image reconstruction etc.

6
4.3 Frequency Spectrum

 In general, the Fourier transform contains

complex terms, and it is customary for display
purposes to work with the magnitude of the
transform (a real quantity), which is called the
Fourier spectrum or the frequency spectrum:

7
(a) A box function, (b) its Fourier transform, and (c) its spectrum.
 4.4 Continuous 2-D Fourier
Transform and Its Inverse
 Let f(t,z) be a continuous function of two
continuous variables, t and z. The two
dimensional, continuous Fourier transform is
given by the expressions

and its inverse is given by

where and are the frequency variables.

8
…cont’d

 When referring to images, t and z are

interpreted to be continuous spatial variables.
 The domain of the variables and defines
the continuous frequency domain.

9
 4.5 Discrete 2-D Fourier
Transform And Its Inverse
 The 2-D discrete Fourier transform (DFT) is
given by:

where f(x,y) is a digital image of size M x N,

= 0, 1, 2 ,… M – 1 and = 0, 1, 2 ,… N - 1

10
…cont’d

 Given the transform , we can obtain

f(x,y) by using the inverse discrete Fourier
transform (IDFT):

for x = 0, 1, 2 ,…, M - 1 and y = 0, 1, 2 , ..., N - 1

11
 Fourier Spectrum And Phase
Angle
 Because the 2-D DFT is complex in general,
it can be expressed in polar form:

where the magnitude

is called the Fourier (or frequency) spectrum,

and

is the phase angle or phase spectrum.

12
…cont’d

 Finally, the power spectrum is defined as

 R and I are the real and imaginary parts of ,

and all computations are carried out for the
discrete variables u = 0, 1, 2 ,…, M - 1 and
v = 0, 1, 2 , . …, N - 1
 Therefore, , , , , and are arrays of
size M x N.
13
 4.6 Frequency Information of
Images
 Frequency in images is the rate of change of
intensity values.
 Thus, a high-frequency image is the one
where the intensity values change quickly
from one pixel to the next.
 On the other hand, a low-frequency image
may be one that is relatively uniform in
brightness or where intensity changes very
slowly.
 Most images contain both high-frequency and
low-frequency components. 14
…cont’d

 Let’s see by an example below

15
…cont’d

 Clearly, in the above image, the zebra pattern

has a high frequency as the intensity changes
very rapidly from white to black.
 While the intensity changes very gradually in
the sky thus it has low frequency.

16
 4.7 Image Enhancement in
Frequency Domain
 Image enhancement is the process of making
images more useful.
 The reason for doing this include:
 Highlighting
 Remove nose from images interesting detail in
images
 Making images more visually appealing

17
Types of enhancement Techniques

 There are two broad categories of image

enhancement techniques:
 Spatial domain techniques – manipulating
image pixels directly
 Frequency domain techniques – manipulating
Fourier transform of images

18
 Frequency Domain Filtering
Fundamentals
 Filtering in the frequency domain consists of
modifying the Fourier transform of an image,
then computing the inverse transform to
obtain the spatial domain representation of
the processed result.
 Thus, given (a padded) digital image, f(x,y),
of size PxQ pixels, the basic filtering equation
in which we are interested has the form:

where is the IDFT, is DFT of the input

image , f(x,y), is a filter transform function19
…cont’d

(which we often call just a filter or filter function),

and g(x,y) is the filtered (output) image.
 Functions F, H, and g are arrays of size PxQ,
the same as the padded input image.
 The product is formed using
elementwise multiplication.
 The filter transfer function modifies the
transform of the input image to yield the
processed output, g(x,y).

20
 Filtering steps in Frequency
Domain
 The process of filtering in the frequency
domain can:
1. Given an input image f(x,y) of size MxN,
obtain the padding sizes P and Q using
P=2M and Q=2N.
2. Form a padded image fp (x,y) of size PxQ
using zero-, mirror-, or replicate padding
3. Multiply fp (x,y) by (-1)x+y to center the Fourier
transform on the PxQ frequency rectangle.
4. Compute the DFT, , of the image from
Step 3. 21
…cont’d

5. Construct a real, symmetric filter transfer

function, , of size PxQ with center at
(P/2, Q/2).
6. Form the product using
elementwise multiplication; that is,
for i = 0, 1, 2, …, M – 1, and k = 0, 1, 2, …, N – 1
7. Obtain the filtered image (of size P × Q) by
computing the IDFT of :

22
…cont’d

8. Obtain the final filtered result, g (x, y), of the

same size as the input image, by extracting
the MxN region from the top, left quadrant of
gp (x,y)

23
Example

Step 1 Step 2 Step 3

Step 4 Step 5 Step 6

Step 7 Step 8
24
Frequency Domain Filtering

 Filtering in the frequency domain consists of

modifying the Fourier transform of an image
and then computing the inverse transform to
obtain the processed result.
 There are two types of Frequency domain
filters:
 Lowpass frequency domain filters and
 Highpass frequency domain filters

25
 Lowpass Frequency domain
Filtering
 Edges and other sharp intensity transitions
(such as noise) in an image contribute
significantly to the high frequency content of
its Fourier transform.
 Hence, smoothing (blurring) is achieved in
the frequency domain by high-frequency
attenuation.
 Attenuating high frequency and passing low
frequency is know as lawpass filtering.
 Three types of lowpass filters: ideal,
Butterworth, and Gaussian. 26
Ideal Lowpass Filters

 A 2-D lowpass filter that passes without

attenuation all frequencies within a circle of
radius from the origin, and “cuts off” all
frequencies outside this circle is called an
ideal lowpass filter (ILPF); It is specified by
the transfer function:

where D0 is a positive constant, and D(u,v) is the

distance between a point (u,v) in the frequency
domain and the center of the PxQ frequency
rectangle; that is, 27
…cont’d

 The name ideal indicates that all frequencies

on or inside a circle of radius D0 are passed
without attenuation, whereas all frequencies
outside the circle are completely attenuated
(filtered out).

Perspective plot of an ideal Function displayed Radial cross section

lowpass-filter transfer as an image 28
function
…cont’d

 For an ILPF cross section, the point of

transition between the values H(u,v) = 1 and
H(u, v) = 0 is called the cutoff frequency.

a b c

d e f

(a) Original image. (b)–(f) Results of filtering using ILPFs with cutoff
frequencies set at radii values 10, 30, 60, 160, and 460
29
Gaussian Lowpass Filters

 Gaussian lowpass filter (GLPF) transfer

functions have the form

where, D(u, v) is the distance from the center of

the PxQ frequency rectangle to any point, (u,v)
contained by the rectangle.
 By letting , we can express the
Gaussian transfer function

where D0 is the cutoff frequency.

30
…cont’d

 When D(u,v) = D0 the GLPF transfer function

is down to 0.607 of its maximum value of 1.0.

Perspective plot of a GLPF Function displayed Radial cross section

transfer function as an image

31
…cont’d

(a) Original image. (b)–(f) Results of filtering using GLPFs

with cutoff frequencies 32
Butterworth Lowpass Filters

 The transfer function of a Butterworth

lowpass filter (BLPF) of order n, with cutoff
frequency at a distance D0 from the center of
the frequency rectangle, is defined as

33
…cont’d

Perspective plot of a BLPF Function displayed Radial cross sections

transfer function as an image. of BLPFs of orders 1
through 4.

34
…cont’d

(a) Original image. (b)–(f) Results of filtering using BLPFs

with cutoff frequencies and n = 2.5
35
Summary

 Ideal, Butterworth, and Gaussian cover the

range from very sharp (ideal) to very smooth
(Gaussian) filtering.
 The shape of a Butterworth filter is controlled
by a parameter called the filter order. For
large values of this parameter, the
Butterworth filter approaches the ideal filter.
 For lower values, the Butterworth filter is
more like a Gaussian filter.
 Thus, the Butterworth filter provides a
transition between two “extremes.” 36
 Highpass Frequency Domain
Filtering
 Because edges and other abrupt changes in
intensities are associated with high-frequency
components, image sharpening can be
achieved in the frequency domain by
highpass filtering, which attenuates low-
frequencies components without disturbing
high-frequencies in the Fourier transform.
 Three types of lowpass filters: ideal,
Butterworth, and Gaussian

37
…cont’d

 Subtracting a lowpass filter transfer function

from 1 yields the corresponding highpass
filter transfer function in the frequency
domain:

where is the transfer function of a

lowpass filter.

38
 Ideal Highpass Filters

 An ideal highpass filter (IHPF) transfer

function is given by

where, as before, D(u,v) is the distance from the

center of the P x Q frequency rectangle

39
…cont’d

 Perspective plot, image, and, radial cross

section of an IHPF transfer function.

40
…cont’d

 Example, image filtered with IHPF transfer

functions using D0 = 60 ( first row) and D0 =
160 (second row).

41
Gaussian Highpass Filters

 The transfer function of a Gaussian highpass

filter (GHPF) transfer function is given by

 Perspective plot, image, and, radial cross

section of an GHPF transfer function.

42
…cont’d

 Example, image filtered with GHPF transfer

functions using D0 = 60 ( first row) and D0 =
160 (second row).

43
Butterworth Highpass Filters

 The transfer function of a Butterworth

highpass filter (BHPF) is

 Perspective plot, image, and, radial cross

section of an BHPF transfer function.

44
…cont’d

 Example, image filtered with GHPF transfer

functions using D0 = 60 in all cases (n = 2 for
the BHPF) ( first row) and D0 = 160 (second
row).

45
…cont’d

 As before, we see that the BHPF transfer

function represents a transition between the
sharpness of the IHPF and the broad
smoothness of the GHPF transfer function.

46
 Chapter Five

Image Restoration and Reconstruction

1
5.1 Introduction

 The principal goal of restoration techniques is

to improve an image.
 Image restoration is an objective process but
image enhancement is largely a subjective
process.
 Restoration recover an image that has been
degraded by using a priori knowledge of the
degradation phenomenon.
 Restoration techniques model the
degradation and apply the inverse process in
order to recover the original image. 2
 5.2 A Model of Image
Degradation/Restoration Process
 Model image degradation as an operator
that, together with an additive noise term,
operates on an input image f(x,y) to produce
a degraded image g(x,y)
 Given g(x,y), some knowledge about , and
some knowledge about the additive noise
term , the objective of restoration is to
obtain an estimate of the original image.
 The more we know about and the closer
will be to f (x, y)
3
Degradation and Restoration
model

4
…cont’d

 If is a linear, position-invariant operator,

then the degraded image is given in the
spatial domain by

where is the spatial representation of the

degradation function.
 The degraded image is given in the
Frequency domain by

5
5.3 Noise Models

 The principal sources of noise in digital

images arise during image acquisition and/or
transmission.
 In acquiring images, light levels and sensor
temperature are major factors affecting the
amount of noise in the resulting image.
 An image transmitted using a wireless
network might be corrupted by lightning or
other atmospheric disturbance.

6
Spatial & Frequency Properties Of
Noise
 Spatial properties define the spatial
characteristics of noise, and whether the
noise is correlated with the image.
 Frequency properties refer to the frequency
content of noise in the Fourier (frequency)
domain.

7
 Noise Probability Density
Functions (PDF)
Gaussian noise model
 The PDF of a Gaussian random variable, z, is
defined by the following expression:

where z represents intensity, is the mean

value of z, and is its standard deviation.
Rayleigh Noise Model
 The PDF of Rayleigh noise is given by

8
…cont’d

 The mean and variance of z when this random

variable is characterized by a Rayleigh PDF
are

and

Erlang (Gamma) Noise Model

 The PDF of Erlang noise is

9
…cont’d

where the parameters are such that a > b , b is a

positive integer, and “!” indicates factorial. The
mean and variance of z are

and

Exponential Noise Model

 The PDF of exponential noise is given by

where a>0
10
…cont’d

 The mean and variance of z are

and

 Note that this PDF is a special case of the

Erlang PDF with b = 1.
Uniform Noise Model
 The PDF of uniform noise is

11
…cont’d

 The mean and variance of z are

and

Salt-and-Pepper Noise Model

 The PDF of salt-and-pepper noise is given by

where V is any integer value in the range

12
0<V<Vk−1.
…cont’d

 The mean of salt-and-pepper noise is given by

and the variance by

13
Some important PDF

14
…cont’d

Test pattern used to illustrate the

characteristics of the PDFs
15
…cont’d

Images and histograms resulting from adding Gaussian, Rayleigh,

and Erlanga noise to the image 16
 5.5 Restoration in the Presence of
Noise
 When an image is degraded only by additive
noise

and

 Spatial filtering is used to estimate f(x,y) [i.e.,

de-noising image g(x,y)] in situations when
only additive random noise is present.

19
 Mean Filters

Arithmetic Mean Filter

 The arithmetic mean filter is the simplest of the
mean filters
 Let Sxy represent the set of coordinates in a
rectangular subimage window (neighborhood)
of size mxn, centered on point (x,y).The
arithmetic mean filter computes the average
value of the corrupted image, g(x,y), in the
area defined by Sxy.
 The value of the restored image fˆ at point
(x,y) is the arithmetic mean computed using
20
the pixels in the region defined by S .
…cont’d

 In other words,

where, r and c are the row and column

coordinates of the pixels contained in the
neighborhood Sxy .
 This operation can be implemented using a
spatial kernel of size mxn in which all
coefficients have value 1/mn. A mean filter
smooths local variations in an image, and
noise is reduced as a result of blurring.
21
…cont’d

Geometric Mean Filter

 An image restored using a geometric mean
filter is given by the expression

where indicates multiplication.

 Here, each restored pixel is given by the
product of all the pixels in the sub-image area,
raised to the power 1/mn.
22
…cont’d

 A geometric mean filter achieves smoothing

comparable to an arithmetic mean filter, but it
tends to lose less image detail in the process.
Harmonic Mean Filter
 The harmonic mean filtering operation is given
by the expression

 The harmonic mean filter works well for salt

noise, but fails for pepper noise.
 It does well also with other types of noise like
Gaussian noise. 23
…cont’d

Contraharmonic Mean Filter

 The contraharmonic mean filter yields a
restored image based on the expression

where Q is called the order of the filter.

 This filter is well suited for reducing or
eliminating the effects of salt-and-pepper
noise.
24
…cont’d

 For positive values of Q, the filter eliminates

pepper noise.
 For negative values of Q, it eliminates salt
noise.
 Note that the contraharmonic filter reduces to
the arithmetic mean filter if Q = 0, and to the
harmonic mean filter if Q = −1

25
Example …
(a) Original image
(b) Image corrupted
by additive
Gaussian noise.
(c) Result of
filtering with
an arithmetic
mean filter of size
3x3 . (a) (b)
(d) Result
of filtering with a
geometric mean
filter of the same
size.

(c) (d)
26
 Order Statistic Filters

 Order-statistic filters are spatial filters whose

response is based on ordering (ranking) the
values of the pixels contained in the
neighborhood encompassed by the filter.
 The ranking result determines the response
of the filter.
Median Filter
 Median filter replaces the value of a pixel by
the median of the intensity levels in a
predefined neighborhood of that pixel:
27
…cont’d

where, as before, Sxy is a subimage

(neighborhood) centered on point (x,y)
 Median filters provide excellent noise-
reduction capabilities
Max and Min Filters
 Max filter is given by

28
…cont’d

 This filter is useful for finding the brightest

points in an image or for eroding dark regions
adjacent to bright areas. It reduce pepper
noise.
Min Filter
 The 0th percentile filter is the min filter:

 This filter is useful for finding the darkest

points in an image or for eroding light regions
adjacent to dark areas.
 It reduces salt noise. 29
…cont’d

Midpoint Filter
 The midpoint filter computes the midpoint
between the maximum and minimum values in
the area encompassed by the filter:

 It works best for randomly distributed noise,

like Gaussian or uniform noise

30
…cont’d

Alpha-Trimmed Mean Filter

 Suppose that we delete the d/2 lowest and the
d/2 highest intensity values of g(r,c) in the
neighborhood Sxy . Let gR (r,c) represent the
remaining mn - d pixels in Sxy .
 A filter formed by averaging these remaining
pixels is called an alpha-trimmed mean filter.
The form of this filter is

where value of d can range from 0 to mn − 1 31

…cont’d

 When d = 0 the alpha-trimmed filter reduces to

the arithmetic mean filter discussed earlier.
 If we choose d = mn − 1, the filter becomes a
median filter.
 For other values of d, the alpha-trimmed filter
is useful in different types of noise, such as a
combination of salt and-pepper and Gaussian
noise.

32
 Adaptive Filters

 The filters discussed so far are applied to an

image without regarding how image
characteristics vary from one point to another.
 Adaptive filters are applied to an image by
considering how image characteristics vary
from one point to another.
 Adaptive filters are capable of performance
superior to that of the filters discussed so far.
 There are two types of adaptive filters:
 Adaptive Local Noise Reduction Filter
 Adaptive median filter
33
 Adaptive, Local Noise Reduction
Filter
 Adaptive filter bases on mean and variance
because they are quantities closely related to
the appearance of an image.
 The mean gives a measure of average
intensity in the region over which the mean is
computed, and the variance gives a measure
of image contrast in that region.
 An adaptive expression for obtaining fˆ(x,y) is
given by

34
…cont’d

, the variance of the noise

, the local average intensity of the pixels in Sxy
, the local variance of the intensities of pixels in
Sxy

35
Adaptive Median Filter

 The adaptive median-filtering algorithm uses

two processing levels, denoted level A and
level B, at each point (x,y) :

36
…cont’d

where

37
 Chapter Six

Image Compression

1
 6.1 Basic Definition of Image
Compression
 Data compression refers to the process of
reducing the amount of data required to
represent a given quantity of information.
 Data are the means by which information is
conveyed.
 Various amounts of data can be used to
represent the same amount of information
 Representations that contain irrelevant or
repeated information are said to contain
redundant data.
2
…cont’d

 Image compression, the art and science of

reducing the amount of data required to
represent an image
 Let b and b′ denote the number of bits in two
representations of the same information, the
relative data redundancy, R, of the
representation with b bits is

3
…cont’d

where C, commonly called the compression

ratio, is defined as

 For instance, if C = 10 (sometimes written

10:1), the larger representation has 10 bits
of data for every 1 bit of data in the smaller
representation. The corresponding relative
data redundancy of the larger
representation is 0.9 (R = 0.9), indicating
that 90% of its data is redundant.
4
6.2 Data Redundancy

 Three forms of data redundancies exist in digital

images
 Coding Redundancy
 Spatial and Temporal Redundancy
 Irrelevant Information.

5
Coding Redundancy

 A code is a system of symbols (like, letters,

numbers, bits) used to represent a body of
information.
 Each piece of information is assigned a
sequence of code symbols, called a code
word.
 The number of symbols in each code word is
its length.
 The 8-bit codes that are used to represent the
intensities in most 2-D intensity arrays contain
more bits than are needed to represent the
6
intensities.
…cont’d

 Let rk in the interval [0,L – 1] is used to

represent the intensities of an M*N image,
and that each rk occurs with probability pr (rk).

where L is the number of intensity values,

and nk is the number of times that the kth
intensity appears in the image.
 If the number of bits used to represent each
value of rk is lr(k), then the average number of
bits required to represent each pixel is
7
…cont’d

 The total number of bits required to represent

an M*N image is M*N*Lavg.
 If the intensities are represented using a
natural m-bit fixed-length code, then the total
number of bits required to represent an M*N
image is M*N*m .

8
 …cont’d

 Example

 Lavg for code 1 is 8 bits.

 The average length of encoded pixels for code 2
Lavg = 0.25*2 + 0.47*1 + 0.25*3 + 0.03*3 = 1.81bits
C = 256*256*8/256*256*1.81 = 4.42
R = 1 – 1/C = 1-1/4.42 = 0.774  Thus, 77.4% of the
data in the original 8-bit 2-D intensity array is
redundant. 9
Spatial and Temporal Redundancy

 Because the pixels of most 2-D intensity

arrays are correlated spatially (i.e., each pixel
is similar to or dependent upon neighboring
pixels), information is unnecessarily
replicated in the representations of the
correlated pixels.
 In a video sequence, temporally correlated
pixels (i.e., those similar to or dependent
upon pixels in nearby frames) also duplicate
information.
10
…cont’d

 To reduce the redundancy associated with

spatially and temporally correlated pixels, a 2-
D intensity array must be transformed into a
more efficient representation.
 For example, run-lengths (128:1) or the
differences between adjacent pixels can be
used.
 In run-length representation, each run-length
pair specifies the start of a new intensity and
the number of consecutive pixels that have
that intensity. 11
Irrelevant Information

 Most 2-D intensity arrays contain information

that is ignored by the human visual system
and/or extraneous to the intended use of the
image.
 It is redundant in the sense that it is not used.

12
…cont’d

(a) coding redundancy, (b) spatial redundancy, (c) irrelevant information

13
 6.3 Elements of Information
Theory
 A key question in image compression is: “
what is a minimum amount of data that is
sufficient to describe an image without losing
information?”
 How do we measure the information content
of an image?
 Average information content of an image (the
intensity source’s entropy) is given by:

units/pixel
14
Example

 It is not possible to code the intensity values

of the imaginary source (and thus the sample
image) with fewer than H̃ bits/pixel.
15
Fidelity Criteria

 Since information may be lost during

compression a means of quantifying the
information loss is desirable.
 Two types of criteria:
 Objective fidelity criteria
 Subjective fidelity criteria
 Objective fidelity criteria: the information loss
is expressed as a function of input image
(original image) and output image
(compressed and decompressed)
16
….cont’d

 The Root-mean-squared (rms) error between

f(x,y) and fˆ(x,y) is

 The Means-Quared Signal-to-Noise ratio of

the output image, denoted SNRms, can be
defined as

17
…cont’d

 Subjective fidelity criteria: measuring image

quality by the subjective evaluation of a group
of viewers and averaging their evaluations.

18
 6.4 General Mechanism of Data
Compression
 An image compression system is composed
of two distinct functional components: an
encoder and a decoder.
 The encoder performs compression, and the
decoder performs the complementary
operation of decompression.
 A codec is a device or program that is
capable of both encoding and decoding.

19
…cont’d

Functional block diagram of a general image

compression system.

20
 The Encoding or Compression
Process
 The encoder is designed to remove the data
redundancies through a series of three
independent operations.
 In the first stage of the encoding process, a
mapper transforms f(x,y) into a format
designed to reduce spatial and temporal
redundancy. This operation generally is
reversible, and may or may not directly
reduce the amount of data required to
represent the image.
 Example, Run-Length Coding 21
…cont’d

 The quantizer reduces the accuracy of the

mapper’s output in accordance with a pre-
established fidelity criterion. The goal is to
remove irrelevant information from the
compressed representation. This operation is
irreversible.
 In the third and final stage of the encoding
process, the symbol coder generates a fixed-
length or variable-length code to represent
the quantizer output, and maps the output in
accordance with the code. 22
…cont’d

 The shortest code words are assigned to the

most frequently occurring quantizer output
values, thus minimizing coding redundancy.
This operation is reversible.

23
 The Decoding or Decompression
Process
 The decoder contains only two components:
a symbol decoder and an inverse mapper.
 They perform, in reverse order, the inverse
operations of the encoder’s symbol encoder
and mapper.
 Because quantization results in irreversible
information loss, an inverse quantizer block is
not included in the general decoder model.

24
6.5 Types of Data Compression

25
Huffman Coding

 When coding the symbols of an information

source individually, Huffman coding yields the
smallest possible number of code symbols
per source symbol.
 The first step in Huffman’s approach is to
create a series of source reductions by
ordering the probabilities of the symbols
under consideration, then combining the
lowest probability symbols into a single
symbol that replaces them in the next source
reduction. 26
…cont’d

 This process is then repeated until a reduced

source with two symbols is reached.

Huffman source reductions.

27
…cont’d

 The second step in Huffman’s procedure is to

code each reduced source, starting with the
smallest source and working back to the
original source. The minimal length binary
code for a two-symbol source are the
symbols 0 and 1
 This operation is then repeated for each
reduced source until the original source is
reached.

28
…cont’d

Huffman code assignment procedure.

 The average length of this code is

and the entropy of the source 2.14 bits/symbol

29
Arithmetic Coding

 In arithmetic coding, a one-to-one

correspondence between source symbols
and code words does not exist.
 Instead, an entire sequence of source
symbols (or message) is assigned a single
arithmetic code word.
 The code word itself defines an interval of
real numbers between 0 and 1

30
…cont’d

 As the number of symbols in the message

increases, the interval used to represent it
becomes smaller, and the number of
information units (say, bits) required to
represent the interval becomes larger.
 Each symbol of the message reduces the
size of the interval in accordance with its
probability of occurrence.
 The output of arithmetic coding is a stream of
bits.
31
…cont’d

 In order to explain the algorithm number will

be shown as decimal, but obviously they are
always binary.

32
Arithmetic coding example

 A five-symbol sequence or message,

a1a2a3a3a4, from a four-symbol source is
coded.
 At the start of the coding process, the
message is assumed to occupy the entire
half-open interval [0, 1).
 This interval is subdivided initially into four
regions based on the probabilities of each
source symbol.

33
…cont’d

Symbol a1, for example, is associated with subinterval

[0, 0.2). Because it is the first symbol of the message
being coded, the message interval is initially narrowed
to [0, 0.2). Thus, [0, 0.2) is expanded to the full height of
the figure, and its end points labeled by the values of
the narrowed range. The narrowed range is then
subdivided in accordance with the original source
symbol probabilities, and the process continues until all
34
the symbols are coded.
Arithmetic coding procedure.

[0.06752, 0.688)  0.068 (a1a2a3a3a4 )

35
Bit Plane Coding

 Bit-plane coding, is based on the concept of

decomposing a multilevel (monochrome or
color) image into a series of binary images
and compressing each binary image
 The intensities of an m-bit monochrome
image can be represented in the form of the
base-2 polynomial

36
…cont’d

 Based on this property, a simple method of

decomposing the image into a collection of
binary images is to separate the m
coefficients of the polynomial into m 1-bit bit
planes.
 The lowest-order bit plane (the plane
corresponding to the least significant bit) is
generated by collecting the a0 bits of each
pixel, while the highest-order bit plane
contains the am-1 bits or coefficients.
37
 …cont’d
 In general, each bit plane is constructed by setting its
pixels equal to the values of the appropriate bits or
polynomial coefficients from each pixel in the original
image.
 Disadvantages of this approach is that small changes
in intensity can have a significant impact on the
complexity of the bit planes.
 E.g., because the most significant bits of the binary
codes for 127(01111111) and 128(10000000) are
different, the highest bit plane will contain a zero-
valued pixel next to a pixel of value 1, creating a 0 to
1 (or 1 to 0) transition at that point. 38
…

 An alternative decomposition approach

(which reduces the effect of small intensity
variations) is to first represent the image by
an m-bit Gray code.
 The m-bit Gray code gm-1… g2 g1 g0 that
corresponds to the base-2 polynomial can be
computed from

39
…cont’d

 This code has the unique property that

successive code words differ in only one bit
position.
 Thus, small changes in intensity are less
likely to affect all m bit planes.
 For instance, when intensity levels 127 and
128 are adjacent, only the highest-order bit
plane will contain a 0 to 1 transition, because
the Gray codes that correspond to 127 and
128 are 01000000 and 11000000,
respectively. 40
 …cont’d

The four most significant binary and Gray-coded bit planes of the image

41
…cont’d

The four least significant binary and Gray-coded bit planes of the image

42
 Chapter Seven

Color Image Processing

1
Color Fundamentals

 Color is the aspect of things that is caused by

different qualities of light being reflected or
emitted by them.
 The characteristics used to distinguish on
color from another are:
 Hue – refers to primary colors - red, green and
blue and secondary colors - magenta (red plus
blue), cyan (green plus blue), and yellow (red
plus green)
 Brightness – the amount of intensity
 Saturation – the degree of purity of hue 2
…cont’d

 Motivation to use color in image processing :

 Color is a powerful descriptor that simplifies
object identification and extraction from a
scene.
 Humans can perceive thousands of color
shades as opposed to only about two dozen
shades of gray.
 Color image processing is divided into two
major areas:
 Full-color processing – image are acquired
and processed in full color 3
…cont’d

 Pseudo-color processing – images are by

nature grayscale and are converted to color
images for visualization purposes.
 In 1666, Isaac newton discovered that when
a beam of sunlight passes through a glass
prism, the emerging beam of light is split into
a spectrum of colors ranging from violet at
one end to red at the other.

4
…cont’d

 Three basic quantities used to describe the

quality of a chromatic light source are:
 Radiance - the total amount of energy that
flows from the light source
 Luminance - a measure of the amount of
energy that an observer perceives from a light
source, and
 Brightness - a subjective descriptor that is
practically unmeasurable.

5
…cont’d

 The human eye sees colors as variable

combinations of primary colors: red (R),
green (G), and blue (B).
 The primary colors can be added together to
produce the secondary colors of light—
magenta (red plus blue), cyan (green plus
blue), and yellow (red plus green).
 Mixing the three primaries, or a secondary
with its opposite primary color, in the right
intensities produces white light.
6
…cont’d

7
…cont’d

 Hue and saturation taken together are called

chromaticity
 A color may be characterized by its
brightness and chromaticity
 The amounts of red, green, and blue needed
to form any particular color are called the
tristimulus values, and are denoted, X, Y, and
Z, respectively.
 A color is then specified by its trichromatic
coefficients, defined as
8
…cont’d

9
 7.2 Color Models (color space or
color system)
 The color model aims to facilitate the
specification of colors in some standard way.
 In digital image processing, commonly used
color models are:
 RGB (Red, Green, Blue)
 CMY (Cyan, Magenta, Yellow)
 CMYK (Cyan, Magenta, Yellow, Black)
 HSI (Hue, Saturation, Intensity)

10
The RGB Model

 Each color appears in its primary colors red,

green and blue.
 This model is based on a Cartesian
coordinate system.
 All color values R, G, and B are normalized in
the range [o, 1]
 We can represent each of R, G, and B from 0
to 255
 Each RGB color image consist of three
component images, one for each primary
11
color.
…cont’d

 The number of bits used to represent each

pixel in RGB space is called the pixel depth.
 24-bit image often referred as full-color image

12
CMY Model

 This model is made up of secondary colors of

light – Cyan, Magenta and Yellow
 Light reflected from a surface coated with
pure cyan does not contain red
 Similarly, pure magenta does not reflect
green, and pure yellow does not reflect blue.
 Therefore,

13
CMYK Model

 CMYK color space is a variation on the CMY

model.
 It adds black
 When equal components of cyan, magenta
and yellow inks are mixed the result is usually
a dark brown not black
 Adding black ink solve problem

14
HSI Model

 When humans view a color object, we

describe it by its hue, saturation, and
brightness.
 Hue is a color attribute that describes a pure
color (pure yellow, orange, or red), whereas
saturation gives a measure of the degree to
which a pure color is diluted by white light.
 Brightness is a subjective descriptor that is
practically impossible to measure.

15
…cont’d

 The HSI (hue, saturation, intensity) color

model, decouples the intensity component
from the color-carrying information (hue and
saturation) in a color image.
 As a result, the HSI model is a useful tool for
developing image processing algorithms

16
 7.3 Basics of Full-Color Image
Processing
 Full-color image processing approaches fall
into two major categories:
 Processing each grayscale component image
individually, then form a composite color
image from the individually processed
components.
 Processing color pixels directly.
 Full-color images have at least three
components, and hence color pixels are
vectors.
17
…cont’d

 For example, in the RGB system, each color

point can be interpreted as a vector extending
from the origin to that point in the RGB
coordinate system
 Let c represent an arbitrary vector in RGB
color space:

 c are the RGB components of a color image

at a point.
18
…cont’d

 The colors of the pixels in an image are a

function of spatial coordinates (x, y) by using
the notation

 For an image of size MxN, there are MN such

vectors,c(x, y), for x = 0, 1, 2 ,…, M – 1 and
y = 0,1, 2 , …, N -1 .
19
…cont’d

 The results of individual color component

processing are not always equivalent to direct
processing in color vector space.
 In this case we must use approaches for
processing the elements of color points
directly.
 We use the terms vectors, points, and voxels
interchangeably when we are referring to
images composed of more than one 2-D
image.
20
…cont’d

 In order for per-component-image and vector-

based processing to be equivalent, two
conditions have to be satisfied:
 first, the process has to be applicable to both
vectors and scalars;
 second, the operation on each component of a
vector (i.e., each voxel) must be independent
of the other components.

21
…cont’d

 Spatial neighborhood processing of grayscale

and full-color images

(a) (b) 22
…cont’d
 In (a), averaging would be done by summing
the intensities of all the pixels in the 2-D
neighborhood, then dividing the result by the
total number of pixels in the neighborhood.
 In (b), averaging would be done by summing
all the voxels in the 3-D neighborhood, then
dividing the result by the total number of
voxels in the neighborhood.
 Each of the three component of the average
voxel is the sum of the pixels in the single
image neighborhood centered on that
location. 23
…cont’d

 But the same result would be obtained if the

averaging were done on the pixels of each
image, independently, and then the sum of
the three values were added for each.
 Thus, spatial neighborhood averaging can be
carried out on a per-component-image or
directly on RGB image voxels.

24
 7.4 Morphological image
processing
 Mathematical morphology a tool used for
extracting image components that are useful
in the representation and description of
region shape, such as boundaries, skeletons,
and the convex hull.
 In image processing, we use morphology with
two types of sets of pixels: objects and
structuring elements (SE’s).
 Objects are defined as sets of foreground
pixels.
25
…cont’d

 Structuring elements can be specified in terms

of both foreground and background pixels.
 Images are rectangular arrays, and sets in
general are of arbitrary shape, applications of
morphology in image processing require that
sets be embedded in rectangular arrays.
 In forming such arrays, we assign a
background value to all pixels that are not
members of object sets.

26
…cont’d

Left: Objects represented as graphical sets.

Center: Objects embedded in a background to
form a graphical image. Right: Object and
background are digitized to form a digital image

27
…cont’d

 Structuring elements are defined in the same

manner

 Difference between the way we represent

digital images and digital structuring
elements:
 There is a border of background pixels
surrounding the objects, while there is none in
the Structuring element.
28
…cont’d

 The reflection of a set (structuring element)

Bˆ about its origin, denoted by, is defined as

 If B is a set of points in 2-D, then Bˆ is the set

of points in B whose (x, y) coordinates have
been replaced by (-x, -y).

29
…cont’d

 Structuring elements and their reflections

about the origin

30
…cont’d

 The translation of a set B by point z = (z1, z2)

is denoted (B)z , is defined as

 If B is a set of pixels in 2-D, then (B)z is the

set of pixels in B whose (x, y)coordinates
have been replaced by (x + z1 , y + z2 )

31
 For morphological image processing, we
need a structuring element.
 It is similar to a mask used in spatial
convolution.
 Morphological operations are defined for two
images.
 The image being processed is the active
image, and the second image is called
(kernel) or (structuring element)

32
…cont’d

 Each structuring element has a prespecified

shape, which is applied as a filter on the
active image.
 The active image can be modified by masking
it with the structuring elements of different
sizes and shapes.
 The basic operations in mathematical
morphology are dilation and erosion. These
two operations can be combined in sequence
to develop other operations, such as opening
and closing. 33
…cont’d

 Dilation operation:
 Given a set A and the structuring element B,
the dilation of A with B is defined as:

 Generally size of B or B^ is smaller than A. If

B^ is placed at the boundary of A, then the
size of A increases to include B^ points. So, all
the points touching the boundary will be
included because of the dilation.
34
…cont’d

 If there is a very small object, say (hole) inside

the object A, then this unfilled hole inside the
object is filled up.
 Small disconnected regions outside the
boundary may be connected by dilation.
 Irregular boundary may also be smoothened
out by dilation. Dilation is translation invariant.

35
…cont’d

Low-resolution text showing broken characters (b)

Structuring element. (c) Dilation of (a) by (b). Broken
segments were joined. 36
…cont’d

 Erosion operation:
 Erosion of A with B is given by:

 In this operation, the structuring element B

should be completely inside the object A, and
that is why the boundary pixels are not
included.
 Two nearly connected regions will be
separated by erosion operation.
37
…cont’d

 Any hole inside the object will be increased

and boundary of an object may be
smoothened by the erosion operation.
 Dilation and erosion are dual operations in
the sense that,

 Dilation expands the components of a set and

erosion shrinks it.
38
…cont’d

 Closing operation:
 The closing of set A by structuring element B,
is defined as

 Closing tends to smooth sections of contours,

it generally fuses narrow breaks and long thin
gulfs, eliminates small holes, and fills gaps in
the contour.

39
…cont’d

(a) Image I, composed of set (object) A, and

background. (b) Structuring element B. (c) Translations
of B such that B does not overlap any part of A. (A is
shown dark for clarity.) (d) Closing of A by B.
40
…cont’d

 Opening Operation
 The opening of set A by structuring element B,
is defined as

 Opening generally smoothes the contour of an

object, breaks narrow isthmuses, and
eliminates thin protrusions.

41
…

Image I, composed of set (object) A and background.

(b) Structuring element, B. (c) Translations of B while
being contained in A. (A is shown dark for clarity.) (d)
Opening of A by B. 42