Professional Documents
Culture Documents
Introduction
Digital Image Processing is an exciting field that evolves rapidly with growing
applications in all fields of science and engineering. The growth of this field is
intimately tied to the development of digital computers and is fuelled by tech-
nological advances in digital imaging and mass storage devices. Digital image
processing has a broad spectrum of applications ranging from remote sensing via
satellites to automated inspection of industrial parts, medical image processing
etcetera.
Image
Digital Image
An image is said to be a digital image if it is in computer readable format so that
it can be stored in a computer (or any other processor) for further processing. The
images acquired using a digital camera or by scanning a picture are examples of
digital images. Thus the images that appear on paper are the printed version of a
digital image.
Definition 1.2 An image f(x,y)is said to be digital image, if its spatial coordi-
nates (x,y) and the amplitude values are all finite and discrete quantities. A dig-
ital image is comprised of a finite number of elements called picture elements or
pixels. Each pixel has a specific location and value.
(a) (b)
1. Binary Images (or) Bi-level Images In a binary image as the name indi-
cates each pixel is comprised of 1 bit. Since the pixel size is one it can take
only two values: either 0 or 1. Pixels with value 0 appear as black and pix-
els with value 1 appear as white in the image. The intermediate shades are
not represented. Figure 1.3 (a) shows a typical example for a binary image.
2. Gray scale Images (or) Gray level Images In a gray scale image each
pixel is comprised of k bits, where k is a positive integer. Pixels can take
a range of value from 0 to 2k −1 and hence can represent shades of gray
starting from black to white. Figure 1.3 (b) shows a typical example for
gray scale image.
3. Color Images Color image is an image in which each pixel is specified
by three values - one each for the red, blue, and green components of the
pixel’s color. Each value is comprised of k bits. Figure 1.3 (c) shows a
typical example for color image.
(a) Binary Image (b) Gray Scale Image (c) Color Image
Digital images require so much storage and computational power that progress
in the field of digital image processing has been dependent on the development of
digital computers and of supporting technologies that include data storage, display
and transmission.
The first computer capable of carrying out meaningful image processing task
appeared in the year 1960. Initially digital image processing techniques were used
for space applications for enhancing images taken by spacecrafts. In the year
1964, computers were used for enhancing images at Jet Propulsion Laboratory
(Pasadena, California).
Figure 1.5 shows the first image of the moon taken by the U.S spacecraft
Ranger 7 on July 31, 1964 at 9:09 a.m. The first digital image processing was per-
formed on this image by the on-board television camera to correct the distortions.
In parallel with space applications, digital image processing techniques began
in the late 1960’s and early 1970s to be used in medical imaging, remote sensing
and astronomy. The invention of Computerized Axial Tomography (CAT) in the
year 1970 is one of the most important event in the application of digital image
processing in medical diagnosis. From the year 1960, the field of image process-
ing has grown vigorously. In addition to applications in medicine and space, ge-
ographers use image processing techniques to study pollution patterns from aerial
and satellite imagery.
In archaeology, image processing techniques are used to restore blurred im-
age that record rare artifacts. In physics and related fields, computer techniques
are used to enhance experimental images obtained using high energy plasmas and
electron microscopy. Apart from these applications, other major fields that used
image processing techniques are biology, nuclear medicine, law enforcement, de-
fense and industry. Typical applications in machine perception that use image pro-
cessing techniques are automatic character recognition, industrial inspection, bio-
metric processing, machine processing of aerial and satellite imagery for weather
prediction and environmental assessment.
The decline in the price of computers and the expansion of networking and
communication bandwidth via the world wide web and the Internet have created
unprecedented opportunities for continued growth of this field.
Knowledge base
Knowledge base refers to the knowledge about the problem domain. The knowl-
edge base is used to guide the operation of each processing module and also to
control the interaction between the modules. As the complexity of the image pro-
cessing task increases the complexity of the knowledge base also increases. For
example, image enhancement for human visual interpretation requires a simple
Digital Image Fundamentals 1.7
knowledge base whereas the knowledge base required for pattern recognition can
be quite complex.
Outputs of these steps are generally images
Image
Restoration Segmentation
attributes
Problem Domain
Image Representation
Enhancement Knowledge Base & Description
Image Object
Acquisition Recognition
Image acquisition
The first step in any image processing system is to acquire the images required
for processing. Image acquisition can be as simple as getting an image that is
already in digital form or digitizing an analog image. Digital cameras which cap-
ture images directly in digital form use either Charge Coupled Devices(CCD) or
Complementary Metal Oxide Semiconductor (CMOS)image sensors. An image
sensor is a 2D array of light sensitive elements that convert photons to electrons.
Image enhancement
Image restoration
Image restoration deals with recovering the original image from the degraded im-
age. Image restoration is subjective in the sense that, restoration techniques tend
1.8 Digital Image Processing
Recognition
Recognition is the process of assigning labels to an object based on its descriptors.
For example, consider the problem at hand is to sort nuts and bolts. Based on the
attributes obtained the objects can be classified as nuts if they have a hole and are
of square shape and objects that of elongated shapes are classified as bolts.
Thus the various processes that can be performed on an image are discussed.
As the complexity of the image processing tasks increases, the number of pro-
cesses required to solve the problem also increases.
1. Image Sensors
2. Specialized Image Processing hardware
3. Computer
4. Image Processing software
5. Mass Storage
6. Image displays.
7. Hard copy devices
8. Networks
Image sensors
Image sensors are physical devices used to acquire digital images. Two elements
are required to acquire digital images. First is a physical device that is sensitive to
the energy radiated by the object to be imaged. The second element is the digitizer,
which is used to convert the output of the physical sensing device into digital form.
For example, in digital video camera, the sensors act as the physical device that
produces an electrical output proportional to light intensity. The digitizer converts
these electrical outputs to digital data.
Specialized image processing hardware
Specialized image processing hardware is called the front-end subsystem and its
most important characteristic is its speed. This unit performs functions that re-
quire fast data through puts that the main computer cannot handle. Specialized
image processing hardware usually consists of the digitizer and Arithmetic Logic
Unit (ALU) that performs primitive operations. For example, Specialized Image
Processing hardware is used to average images as they are digitized for the pur-
pose of noise reduction.
Computer
Computer in an image processing system is a general purpose computer which can
range from a PC to a super computer. In certain dedicated applications, Computers
are customized to achieve a required level of performance.
Mass storage
When dealing with thousands of images, providing adequate storage in an im-
age processing system is a big challenge. Digital storage for image processing
applications falls into three principal categories:
1. Short term storage
2. Online storage
3. Archival storage.
1. Short term storage : Short term storage devices are used to access images
at the time of processing. One method of providing short term storage is
computer memory and the second method is using frame buffers. Frame
buffers are specialized boards that can store one or more images that can
be accessed rapidly at video rates (e.g.30 images/sec). Frame buffers allow
virtual instantaneous image zoom, image scroll (Vertical shifts) and image
pan (horizontal shifts). Frame buffers usually are placed in the specialized
image processing hardware unit.
2. Online storage : Online storage devices are used to store images that re-
quire frequent and fast recall. Magnetic disks or optical-media storage de-
vices are used as online storage devices.
3. Archival storage : Archival storage devices are used for massive storage
requirements which are accessed rarely Magnetic tapes and optical disks
housed in ‘jukeboxes’ are the usual media for archival applications.
Image displays
Image displays are the physical devices used to display the final output of the
image processing system. Computer and color television monitors are the most
commonly used display devices. Monitors are driven by the outputs of image
and graphics display cards that are integral part of the computer system. Stereo
displays are special display devices that are implemented in the form of headgear
containing two small displays embedded in goggles worn by the user.
Hard copy
Hard copy devices are used for recording images. Most commonly used devices
for recording are laser printer, film cameras, heat sensitive devices, inkjet de-
vices and digital units such as optical and CD-ROM disks. Film provides the
highest possible resolution but paper is the most commonly used medium. For
presentations, images are displayed on transparent films or projected using a LCD
projector.
1.12 Digital Image Processing
Networking
Networking is used to share the large amount of data involved in image process-
ing applications. In dedicated networks transmission bandwidth is not a problem
but for remote communications via internet, transmission bandwidth is the key
consideration. Evolution of optical fibers and other broad band technologies have
greatly influenced the transmission process and nowadays transmission of mass
data via internet is as efficient as via dedicated networks.
Conjunctiva
Vitreous humor
Zonule fibers
Iris
Fovea
Cornea
Visual axis Macula
lutea
Retina
Figure 1.8 shows a simplified horizontal cross section of the human eye. The
eye is nearly a sphere, with an average diameter of approximately 20mm.
Three membranes enclose the eye:
1. The cornea and sclera outer cover;
2. The choroid; and
3. The retina
Cornea and Sclera
The cornea is a tough, transparent tissue that covers the anterior surface of the
eye. Continuous with the cornea, the sclera is an opaque membrane that encloses
the remainder of the optic globe. Sclera forms the white part of the eye.
Choroid
The choroid lies directly below the sclera. This membrane contains a network of
blood vessels that serve as the major source of nutrition to the eye. Even superfi-
cial injury to the choroid can lead to severe eye damage as a result of inflammation
that restricts blood flow. The choroid coat is heavily pigmented and hence helps to
reduce the amount of extraneous light entering the eye and the backscatter within
the optical globe.
At its anterior extreme, the choroid is divided into the ciliary body and the
iris diaphragm. The iris diaphragm contracts or expands to control the amount of
light that enters the eye.
1.14 Digital Image Processing
180,000
Blind spot Cones
Rods
2
No. of rods or cones per mm
135,000
90,000
45,000
Figure 1.9 shows the distribution of rods and cones for a cross section of
the right eye. Except for the blind spot, the distribution of receptors is radially
symmetric about the fovea. Receptor density is measured in degrees from visual
axis.
Cones are most dense in the center area of the fovea which is the reason for
photopic vision. Rods increase in density from the center out to approximately 20 ◦
off axis and then decrease in density out to the extreme periphery of the retina.
Note No.of cones in the region of highest acuity in the eye is about 3,37,000
elements.
1.4.3 Image Formation in the eye
The principal difference between the lens of the eye and an ordinary optical
lens is that the former is flexible. The shape of the lens is controlled by tension in
the fibers of the ciliary body. To focus on distant objects, the controlling muscles
cause the lens to be relatively flattened. Similarly, these muscles allow the lens to
become thicker in order to focus on objects near the eye.
1.16 Digital Image Processing
The distance between the center of the lens and the retina is called the focal
length. Focal length varies from approximately 17mm to about 14mm, as the
refractive power of the lens increases from its minimum to its maximum. When
the eye focuses on far away objects the lens exhibits its lowest refractive power.
When the eye focuses on a nearby object, the lens is most strongly refractive.
Example
In the Figure 1.10, the observer is looking at a tree 15m high at a distance of
100m. If the height of the retinal image is h (in mm), then
15 h
=
100 17
15 × 17
⇒ h= = 2.5mm
100
The retinal image is reflected primarily in the area of the fovea. Perception then
takes place by the relative excitation of light receptors, which transform radiant
energy into electrical impulses that are ultimately decoded by the brain.
1.4.4 Brightness
If the light is achromatic (void of color), then its only attribute is its intensity.
Brightness is a subjective description that is practically impossible to measure. It
embodies the achromatic notion of intensity.
Subjective Brightness
It is the intensity as perceived by Human Visual System (HVS). It is a loga-
rithmic function of the light intensity incident on the eye.
Simultaneous Contrast
This phenomenon is related to the fact that a region’s perceived brightness does
not depend simply on its intensity. All the inner squares in the Figure 1.13 have
exactly the same intensity; however, they appear to become darker as the back-
ground gets lighter.
1.18 Digital Image Processing
A more familiar example is a piece of paper that seems white while lying on
a desk, but can appear totally black when used to shield the eyes while looking
directly at a bright sky.
Optical illusions
Optical illusion refers to a human perception phenomena, in which the eye fills in
non existing information (or) wrongly perceives geometrical properties of objects.
Optical illusions are a characteristic of the human visual system that is not fully
understood. Figure 1.14 shows examples of optical illusions.
In Figure 1.14 (a) a floating white triangle which doesn’t exist is seen. In
Figure (b), the centre circle on the left appears smaller than that on the right, but
they are in fact of the same size. In Figure (c), horizontal lines that are actually
parallel appear to be sloped due to the presence of the black blocks.
Energy
Filter
Sensing material
Power in
Film
Sensor Rotation
Linear motion
Fig 1.16 Generation of 2-D image using a single sensor
1.20 Digital Image Processing
screw that provides motion in the perpendicular direction. Since mechanical mo-
tion can be controlled with high precision, this method is an inexpensive way to
obtain high-resolution images.
Microdensitometers
Microdensitometer is mechanical digitizer used for acquiring 2-D image. Mi-
crodensitometers use a flat bed, with the sensor moving in two linear directions.
These types of mechanical digitizers sometimes are referred to as microdensito-
meters.
Image Acquisition Using Sensor Strips (Line Sensor)
Imaged area
Linear motion
Sensor strip
A rotating X-ray source provides illumination and the portion of the sensors op-
posite to the source collect the X-ray energy that pass through the object (the
sensors obviously have to be sensitive to X-ray energy). The output of the sensors
is processed by reconstruction algorithms to transform the sensed data into mean-
ingful cross-sectional images. A 3-D digital volume consisting of stacked images
is generated as the object is moved in a direction perpendicular to the sensor ring.
Image
reconstruction
Cross-sectional images
of 3-Dobject
3D Object
X-ray source
n
otio
rm
Lin ea
Sensor ring
Applications
1. Medical and industrial Computerized Axial Tomography (CAT)
2. Magnetic Resonance Imaging (MRI) and
3. Positron Emission Tomography (PET) imaging.
Vidicon is the most popular camera tube used in the television industry. The
Vidicon came into general use in the early 50’s and gained immediate popularity
because of its small size and ease of operation. Vidicon is a storage type camera
tube in which a charge density pattern is formed by the image scene radiation
on a photoconductive surface which is then scanned by a beam of low velocity
electrons.
1. Focusing Coil
2. Horizontal and Vertical Deflection Coils
3. Alignment Coils
Grid no.1
Target 0 to 100V
Light
image
Glass Cathode, 0V
face plate
Grid no.3(beam focus)
275 to 300 V
Grid no.4 Focusing
(decelerator) coil Horizontal and vertical
275 V deflecting coils
Operation
In darkness, the photo-layer has a thickness of about 0.0001cm and behaves as an
insulator and thus the electrons are unable to reach the positively charged metal
film. Therefore a layer of electrons is deposited on the photoconductive layer
to balance the positive charge on the metal coating. Thus after a complete scan
of the target by electron beam, the photoconductor appears as a capacitor with a
positively charged plate on one side and a layer of electrons on the other side.
When light strikes a small area of the photoconductor electrons in that area
start to penetrate and the charge in that area is neutralized. Thus with an optical
image on the target, each point on the gun side of the photo-layer assumes a
potential with respect to the dc supply and its resistance. A pattern of positive
potentials appears on the gun side of the photo-layer, producing a charge image
corresponding to the incident optical image.
Storage action
Though light from the scene falls continuously from the target, each element of
the photo coating is scanned at intervals equal to the frame time. This results
in storage action and net change in resistance depends on the time between two
successive scanning and intensity of incident light.
Signal current
As the electron beam scans the target plate it replaces the lost electrons and re-
stores uniform surface charge. This causes a current flow in the signal electrode
circuit producing a varying voltage across the load resistance R L . The amplitude
1.24 Digital Image Processing
of current and the consequent output voltage across R L are directly proportional
to the light intensity variations on the scene. Note that, since, a large current
would cause a higher voltage drop across RL , the output voltage is most negative
for white areas. The video output voltage developed across the load resistance (50
K − ohms) is adequate and does not need any image or signal multiplication.The
output signal is further amplified by conventional amplifiers before it leaves the
camera unit. This makes the vidicon a much simpler picture tube.
Applications
1. Earlier types of vidicons were used only where there was no fast movement,
because of inherent lag. These applications included slides, pictures, closed
circuit TV etc.
2. The present day improved vidicon finds wide applications in education,
medicine, industry, aerospace and oceanography.
Kodak Compact
Memory
Aperture Flash Card
Zoom IR blocking and 16GB Colour LCD
Lens and Shutter anti - aliasing
filter
A/D Camera
ASIC
PC interface
Imager
Focus
Zoom
DRAM
Buffer
Micro
Flash Processor
Optical
View finder Firmware
Memory
focuses light from the scene onto the CCD or CMOS sensor. The analog sig-
nal from the sensor is amplified and converted to digital form, normally using
a 10 or 12-bit analog to digital (A/D) converter. When the user fully depresses
the shutter button, an image is captured and stored in DRAM. In order to create
the final high-resolution digital image, the digital data is processed by a camera
ASIC (Application Specific IC) digital image processor, or by a high-performance
microprocessor, sometimes supported by a digital signal processor (DSP).
De-mosaicing
Since most digital cameras use a single color image sensor that provides only one
color value for each light sensitive cell, interpolation is used to “fill in” the “miss-
ing” color values for each pixel. This is called “de-mosaicing.” De-mosaicing
provides a full-color, but not yet perfect, image.
White balancing
To get a good color image, the camera must also provide white balancing to com-
pensate for spectral variations in the illumination. While both daylight and indoor
lighting provide “white” light, daylight actually has more energy in the blue por-
tion of the light spectrum and indoor lighting provides more energy in the red por-
tion of the spectrum. An image processing algorithm is used to analyze the scene
and adjust the red and blue signal strengths to match the green signal strength in
white and neutral areas of the picture. Even after white balancing, the color image
is not yet perfect.
1.26 Digital Image Processing
Color correction
The RGB spectral sensitivities of the image sensor do not perfectly match the
way (HVS) see colors. As a result, the color image typically appears desaturated,
muting bright colors like red and blue. Color correction digitally compensates for
this, improving the color reproduction, and also transforming the digital image
into the output color space. The output color space used for most digital cameras
today is RGB, which is a color space that is designed to be ready for display on a
typical computer monitor.
Sharpening
The next image processing step is sharpening. The anti-aliasing filter and other
lens components slightly soften the image captured by the color image sensor. The
picture may also be slightly blurred by the display monitor or printer. Therefore,
digital cameras usually include digital image processing to sharpen the image ap-
plying adaptive spatial filters, which locate and emphasize fine edge details in the
image. In some cases, a photographer may prefer different sharpness levels, for
example, depending on whether the photo is a portrait or a nature shot. To provide
this creative control, some digital cameras offer a variety of different sharpness
settings such as “soft,” “standard” and “sharp.”
Finally the compressed image data is stored in a JPEG image file usually ac-
cording to the Exit format (Exit stands for Exchangeable Image File).
In addition to the compressed JPEG image the file contains meta-data such as:
1. Camera manufacturer and model number
2. Exact date and time the picture was taken
3. Focal length, subject distance, lens fnr, and shutter speed used to take the
picture
4. Ambient light level, flash setting
5. User-selected camera settings (sharpness level, quality level, etc.) and
6. a “thumbnail image”.
Certain cameras even allow short “sound bites”
Recording Color
There are several ways for recording the three colors in a digital camera. The
highest quality cameras use three separate sensors, each with a different filter
over it. Light is directed to the different sensors by placing a beam splitter in the
camera. The advantage of this method is that the camera records each of the three
colours at each pixel location. Unfortunately, cameras using this method tend to
be bulky and expensive.
Digital Image Fundamentals 1.27
The second method is to rotate a series of red, blue and green filters in front
of a single sensor. The sensor records the three separate images in rapid succes-
sion. Since the three images are not taken precisely at the same moment, it is not
practical in the case of candid photography or hand held cameras.
A more economical and practical way to record the three primary colours from
a single image is to permanently place a filter over each individual photo site. By
breaking up the sensor into a variety of red, blue and green pixels, it is possible
to get enough information about the general vicinity of each sensor in order to
enable us to make accurate guesses about the true colour at that location. This
process of looking at the other pixels in the neighborhood of a sensor and making
a guess is called interpolation.
Figure 1.24 shows the block diagram to convert an analog image to digital im-
age. The basic idea behind sampling and quantization is illustrated in Figure. 1.25.
Figure 1.25 (a) shows a continuous image, f (x, y), that is to be converted to dig-
ital form. An image may be continuous with respect to the x- and y-coordinates,
and also in amplitude. To convert it to digital form, we have to sample the function
in both coordinates and in amplitude.
• Digitizing the coordinate values is called sampling.
• Digitizing the amplitude values is called quantization.
Sampling
The amplitude along the line segment AB is plotted to obtain the one-dimensional
function shown in Figure 1.25(b). The random variations are due to image noise.
To sample this function, we take equally spaced samples along line AB, as shown
in Figure 1.25(c).The location of each sample is given by a vertical tick mark in
the bottom part of the figure. The samples are shown as small white squares su-
perimposed on the function. The set of these discrete locations gives the sampled
function of the line segment AB.
Digital Image Fundamentals 1.29
A B
A B
(a) (b)
A B A B
Sampling Quantization
(c) (d)
Generating a digital image. (a) Continuous Image. (b) A scan line from A to B is the
continuous image used to illustrate the concepts of sampling and quantization.
(c) Sampling and quantization.(d) Digital scan line.l
Quantisation
In order to form a digital function, the gray-level values also must be converted
(quantized) into discrete quantities. The amplitude (intensity) value of every sam-
ple is to be quantized. The gray level scale is divided into finite number of discrete
levels. In this example the gray-level scale divided into eight discrete levels, rang-
ing from black to white as shown in Figure 1.25(c). One of the eight discrete gray
level value is assigned to each sampled value based on the vertical proximity of
the sample amplitude value to the discrete gray level value in the scale.
1.30 Digital Image Processing
The digital samples resulting from both sampling and quantization are shown
in Figure 1.25(d).
End of this procedure results in digitization of a line segment AB in the image.
In order to obtain the complete 2-D digital image the procedure is to be carried out
line by line starting at the top of the image until the last line segment is digitized.
When a sensing array is used for image acquisition, there is no motion and the
number of sensors in the array determines the limitations of sampling. Figure
1.26 illustrates this concept. Figure 1.26(a) shows a continuous image projected
onto the plane of an array sensor. Figure 1.26(b) shows the image after sampling
and quantization. Clearly, the quality of a digital image is determined to a large
degree by the number of samples and discrete gray levels used in sampling and
quantization.
The right side of this equation is by definition a digital image. Each element of
this matrix array is called an image element, picture element, pixel, or pel.
The second method is to use a more traditional matrix notation as shown below
to denote a digital image and its elements
a0,0 a0,1 · · · a0,N −1
a1,0 a1,1 . . . a1,N −1
A= . .. ..
.. . .
aM −1,0 aM −1.1 · · · aM −1,N −1
The values M and N should be positive integers. The number of gray levels in
the images is represented using L and it should be typically a integer power of 2
as shown in equation, L = 2k .
Digital Image Fundamentals 1.31
Thus the gray level values are equally spaced integers that lie in the interval
[0, L − 1]
Illumination Source
F(x1,x2)
x2
xy0
x1
-xy0 xx0
xx0 x1
-xy0
-xy0
x2
(a) Fourier transform of a (b) Its region of support
band limited function
Fig 1.28 Band Limited Function
provided the sampling rate is greater than the Nyquist rate, that is
1 1
= Us > 2U0 , = Vs > 2V0
∆x ∆y
If the sampling frequencies Us , Vs are greater than twice the bandwidths, that is
Us > 2U0 , Vs > 2V0 then F (U, V ) can be recovered by a Low Pass Filter with
frequency response,
(
1
(Us ,Vs ) (U, V ) ∈ R,
H (U, V ) = (1.1)
0, otherwise
U*
Quantizer
rL output
U Quantizer U*
rk
U
t1 t2 tk tL+1
r2 Quantizer
error
r1
" #ti+1
(u − ri )3
∆= pu (u) f or1 ≤ i ≤ L (1.4)
3
ti
1 1
∆= (ti+1 − ri )3 P u (ti+1 ) − (ti − ri )3 P u (ti ) (1.5)
3 3
To minimize the mean square error, differentiating ∆ with respect to t k and equat-
ing it to zero yields
rk + rk−1
tk = (1.6)
2
1.34 Digital Image Processing
From Eq w.k.t,
R tk+1
t uP u (u) du
rk = Rktk+1 (1.9)
tk P u (u) du
Properties
The optimal uniform quantizer has the following properties
1. The quantizer output is an unbiased estimate of the input,
i.e., E[u∗] = E[u]
2. The quantization error is orthogonal to the quantizer output,
i.e., E[(u − u∗)u] = 0
Example 1.1 Calculate the storage space required to store an 8-bit image of size
256 × 256 pixels.
Solution:
Given M = N = 256 k = 8 bits
1.36 Digital Image Processing
32 × 32
64 × 64
128 × 128
256 × 256
512 × 512
Checkerboard Pattern
It refers to the smallest discernible change in gray level and it depends on the
number of gray levels. Quantization determines the number of gray levels in an
image. Quantization is the principal factor determining the intensity resolution of
an image. The most commonly used resolution is 28 that is 256 gray level.
False Contouring
Use of insufficient number of gray levels in smooth areas of a digital image causes
False Contouring. When groups of neighboring pixels are quantized to the same
value, regions of constant gray levels are formed, whose boundaries are called
contours as shown in Figure 1.34.
Digital Image Fundamentals 1.39
Fig 1.34 (a) Image with k = 4, 16 gray levels (b) Image with k = 3, 8 gray levels (c) Image with
k = 2, 4 gray levels (d) Image with k = 1, 2 gray levels
Uniform quantization of images requires about 256 gray levels (i.e., 8-bits).
Contouring effects start becoming visible at or below 6-bits/pixel. This effect of
false contouring can be held below the level of visibility by using two methods,
(i) Contrast Quantization (ii) Pseudorandom Quantization
Contrast Quantization
-1
Luminance f() C f ()
MMSE
Luminance to Contrast to
Quantizer
contrast Luminance
1 1
f(x,y) g(x,y) Quantizer g (x,y) f (x,y)
å å to display
K bits +
+
+ -
h(x,y)
Uniform
Pseudorandom
Noise
tization. This pseudorandom noise is also called dither. To display the image, the
same pseudorandom sequence is subtracted from the quantizer output. The effect
is that in the regions of low luminance gradients (i.e., contours), the input noise
causes pixels to go above or below the original decision level, thereby breaking
contours. Figure 1.36 shows the block diagram of a Pseudorandom Quantizer.
1.7 Basic Relationship between pixels
Pixels are the basic elements which form the image and their proper manipulation
gives rise to different appearances. The relationships between pixels are used for
image manipulations. In this section we consider several important relationships
between pixels in digital image. As mentioned before an image is denoted by
f (x, y), when referring to a particular pixel, we use lowercase letters p & q.
shown in Figure 1.37. This set of pixels are called 4-neighbors of p denoted by
N4 (p). In N4 (p), each pixel has a unit distance from (x, y) and some of the neigh-
bors of p lie outside the digital image if (x, y) is on the border of the image.
Diagonal Neighbors
The four diagonal neighbors of p have coordinates (x+1, y+1), (x+1, y−1),
(x − 1, y + 1), (x − 1, y − 1) as shown in Figure 1.38 and are denoted by N D (p).
8-Neighbors
Diagonal neighbors together with the 4-neighbors are called 8-neighbors of
p, denoted by N8 (p). Some of the points in ND (p) and N8 (p) fall outside the
image if f (x, y) is on the border of the image. Eight neighbors are at non-uniform
distance from P . Effectively we have N8 (p) = N4 (p) + ND (p). Figure 1.39
shows the eight neighbors of p.
1.7.2 Adjacency
Definition 1.3 Two pixels that are neighbors and have the same gray level (or)
some other specified similarity criterion are called adjacent pixels. Let V be the
set of gray level values used to define adjacency. In a binary image V = 1, if we
are referring to adjacency of pixels with value 1. In a gray scale image the idea
is same but set V typically contains more elements.
Types of Adjacency
There are three different types of adjacency they are
1.42 Digital Image Processing
0 1 1 0 1 1 0 1 1
0 1 0 0 1 0 0 1 0
0 0 1 0 0 1 0 0 1
Fig 1.40 (a) Arrangement of Pixels, (b) Pixels that are 8-adjacent to centre pixel, (c) m-adjacent
• 4 - Adjacency
• 8 - Adjacency
• m - Adjacency (mixed adjacency)
An example of adjacency
• 4 - Adjacency
Two pixels p and q with values from ‘V ’ are adjacent if ‘q’ is in the set
N4 (p).
• 8 - Adjacency
Two pixels p and q with values from ‘V ’ are 8 - adjacent if ‘q 0 is in the set
of N8 (p).
• m - Adjacency
Mixed adjacency is the modification of 8 adjacency and it is introduced to
eliminate the multi-path connection that often arise when 8 - adjacency is
used. Two pixels ‘p’ and ‘q’ with values from ‘V ’ are m - adjacent,
1.7.3 Connectivity
Connectivity between pixel is a fundamental concept that simplifies the definition
of digital image concepts such as regions and boundaries. To establish if two
pixels are connected it must be determined if they are neighbors and if their gray
level satisfy the specified criterion of similarity. Connectivity between pixels is
used for establishing boundary of an object and components of region in an image.
If the pixels are adjacent in some sense and if their gray level are equal, then we
can connect the pixels.
Digital Image Fundamentals 1.43
Types of Connectivity
There are three types of connectivity they are
1. 4 - Connectivity
2. 8 - Connectivity
3. m - Connectivity
• 4 - Connectivity
Two pixels p and q with values from ‘V ’ are 4 connected if ‘q’ is in the set
N4 (p).
• 8 - Connectivity
Two pixels p and q with valued from ‘V ’ are 8 - connected if q is in the set
of N8 (p).
• m - Connectivity
Two pixels p and q with valued from ‘V ’ are 8 - connected if
(i) q is in the set of N4 (p) or
(ii) q is in ND (p) and the set N4 (p) ∩ N4 (q) is empty
1.7.4 Path or Curve
A digital path or curve from pixel ‘p’ to pixel ‘q’ is a sequence of distinct pixels.
If ‘p’ = ‘q’ then the path is closed path. A path between two pixels ‘p’ and ‘q’
with co-ordinates (x0 , y0 ) and (xn , yn ) respectively is a sequence of pixels with
co-ordinates (x0 , y0 ), (x1 , y1 ), (x2 , y2 ) · · · (xn , yn ). where (xi , yi ) is adjacent to
(xi + 1, yi + 1) for 1 ≤ i ≤ n.
Eg:
15 20 32
V = {20,32,40,50,}
40 50 70
C = {20.32.40}
100 45 60
If ‘p’ and ‘q’ are the pixels of an image of subset ‘S’ then ‘p’ is connected to
‘q’ in ‘s’, if there exists a path from ‘p’ to ‘q’ consisting entirely of pixels in S.
For any pixels ‘p’ in ‘S’, the set of pixels that are connected to it in ‘s’ is called a
connected component of S.
In the above Figure 1.41, consider the pixel arrangement with the set
V = {20, 32, 40, 50}. The center pixel 50, has the connected component C =
{20, 32, 40}.
1.44 Digital Image Processing
1.7.6 Edge
An edge is a local concept that is based on a measure of gray level discontinuity
of a point. Edges are formed by the pixels with derivative value that is higher than
a present threshold. Edge points can be linked to form edge segments and edge
segments are linked to form boundaries.
Conditions
Consider three pixels p, q and z with co-ordinates (x, y)(s, t) and (v, w) respec-
tively. Let D be the distance function which satisfies the following conditions,
• D(p, q) ≥ 0, [D(p, q) = 0; if p = q]
• D(p, q) = D(q, p) and
• D(p, z) ≤ D(p, q) + D(q, z)
Euclidean distance
The Euclidean distance between p & q is defined as,
1/2
De (p, q) = (x − s)2 + (y − t)2 (1.19)
p
(Or) De (p, q) = (x − s)2 + (y − t)2 (1.20)
Digital Image Fundamentals 1.45
The city block distance between two pixels p & q is defined as.
D4 (p, q) = |x − s| + |y − t| (1.21)
In this case, the pixels having a D4 distance from (x, y) form a diamond centered
at (x, y).
Example
The pixels with D4 distance ≤ 2 from (x, y) ie, the center point form the
following contours of constant distance.
2
2 1 2
2 1 0 1 2
2 1 2
2
In this case, the pixels with D8 distance from (x, y) form a square centered at
(x, y).
Eg: The pixels with D8 distance ≤ 2 from (x, y) i.e., the center point form
the following contours of constant distance
2 2 2 2 2
2 1 1 1 2
2 1 0 1 2
2 1 1 1 2
2 2 2 2 2
As observed from the figure the fields of application of digital image pro-
cessing is so varied that some form of organization is required to understand
the breadth of the field. Thus the applications are categorized according to their
source of energy.
2. Acoustic
3. Ultrasonic
4. Electronic and
5. Computers
Photographs, X-ray images, RADAR images and other most familiar images
are generated based on the radiation from the EM spectrum. EM waves can be
conceptualized as a stream of charged particles that travel at the speed of light.
Each charged particle contains a quantum of energy called photon. Figure 1.45
shows the Electromagnetic spectrum obtained when arranged according to the
energy per photon. The EM spectrum ranges from gamma rays (highest energy)
at one end to radio waves (lowest energy) at the other end.
Computer Imaging
Synthetic images are created using computer for artistic effects in films and
other multimedia applications. 3-D modeling is a more structured approach to
1.48 Digital Image Processing
image generation by computer and is used more dominantly in medical field. This
is an area that provides an important intersection between image processing and
computer graphics and is the basis for many 3-D visualization systems (e.g., flight
simulators). Figure 1.46 shows examples of computer - generated images. Since
the original object is created in 3-D, images can be generated in any perspective
from plane projections of the 3-D volume. Images of this type can be used for
medical training and for a host of other applications, such as criminal forensics
and special effects.
Table 1.2 summarizes the various applications of digital image processing
based on the source of energy used.
Table-1.2 Applications of Digital Image Processing
S.No. Source of Field of Application
Energy
1 Gamma Nuclear medicine, Astronomical observations,
Rays Positron Emission Tomography(PET).
2 X- rays Medical Diagnostics, Angiography, Computerized
Axial Tomography(CAT), Industrial Inspection, As-
tronomy
3 Ultraviolet Lithography, industrial inspection, microscopy,
Band LASERs, biological imaging, astronomical observa-
tions and Fluorescence microscopy
4 Visible and Remote sensing, photography, videography,
IR band light microscopy, multi spectral imaging (satel-
lites),automated visual inspection of manufactured
goods, traffic monitoring and surveillance
5 Microwave RADAR
band
6 Radio band Medicine (MRI) and Astronomy
7 Acoustics Geological exploration, industry and medicine
8 Ultrasound Medical diagnostics (Obstetrics)
9 Electronics Electron Microscope, Transmission Electron Micro-
scope(TEM), Scanning Electron Microscope(SEM)
10 Computer Synthetic images, artistic images,3-D modeling and
visualization
The use of color in image processing is motivated by two principal factors. First,
color is a powerful descriptor that often simplifies object identification and ex-
traction from a scene. Second, humans can discern thousands of color shades and
intensities, compared to about only two dozen shades of gray. This second factor
is particularly important in manual image analysis.
Color image processing is divided into two major areas: full color and pseudo
color processing. In the first category, the images in question typically are ac-
quired with a full-color sensor, such as a color T.V. camera or color scanner.
In the second category the problem is one of assigning a color to a particular
monochrome intensity or range of intensities until recently, most digital color im-
age processing was done at the pseudo color level. However, in the past decade,
color sensors and hardware for processing color images have become available at
reasonable prices. The result is that full - color image processing techniques are
now used in a broad range of applications, including publishing, visualization and
the internet.
d
Infra re
Red
Orange
Yellow
Green
White light Blue
Violet
Optical Prism
Ultra violet
Fig 1.47 Color Spectrum
In 1666, Sir Isaac Newton discovered that when a beam of sunlight passes
through a glass prism, the emerging beam of light is not white but consists of a
continuous spectrum of colors ranging from violet at one end to red at the other as
shown in Figure 1.47. When viewed in full color, no color in the spectrum ends
abruptly, but rather each color blends smoothly into the next.
Basically the colors that humans and some other animals perceive in an object
are determined by the nature of the light reflected from the object. As illustrated in
Figure 1.48, visible light is composed of a relatively narrow band of frequencies
in the electromagnetic spectrum. A body that reflects light that is balanced in all
visible wavelengths appears white to the observer. However, a body that favors,
reflectance in a limited range of the visible spectrum exhibits some shades of
color. For example, green objects reflect light with wavelengths primarily in the
500 to 570nm range while absorbing most of the energy at other wavelengths.
1.50 Digital Image Processing
Fig 1.48 Wave Length Comprising the Visible Range of Electromagnetic Spectrum
Yellowish Green
Green
Red
Blue Green
Blue
Orange
Orange
Raddish
Yellow
Bluish purple
Purplish blue
Fig 1.49
intensities produces white light. This result is shown in Figure 1.50 which also il-
lustrates the three primary colors and their combinations to produce the secondary
colors.
Green
Yellow
White Black
Blue
Cyan
Magenta Magenta Blue
Red
Differentiating between the primary colors of light and the primary colors of
pigments or colorants is important. A primary color is defined as one that absorbs
a primary color of light and reflects the other two colors. Therefore, the primary
colors of pigments are magenta, cyan, and yellow, and the secondary colors are
red, green and blue. These colors are shown in Figure 1.50. A proper combination
of the three pigment primaries, or a secondary with its opposite primary, produces
black.
1.52 Digital Image Processing
In terms of digital image processing, the hardware -oriented models most com-
monly used in practice are the RGB (red green blue) model for color monitors
and a broad class of color video cameras, the CMY (cyan, magenta, yellow) and
CMYK (cyan, magenta, yellow, black) models for color printing, and the HSI
(hue, saturation, intensity) model, which corresponds closely with the way hu-
mans describe and interpret color. The HSI model also has the advantage that it
decouples the color and gray-scale information in an image, making it suitable for
many of the gray-scale techniques.
normalized so that the cube shown in Figure 1.51 is the unit cube. That is, all
values of R,G, and B are assumed to be in the range [0,1].
(0,01)
Blue
Cyan
Magenta White
Points along the main diagonal have gray values, from black at the origin to
white at point (1,1,1).
Images represented in the RGB color model consist of three component im-
ages, one for each primary color. When fed into an RGB monitor, these three
images combine on the screen to produce a composite color image, as explained.
The number of bits used to represent each pixel in RGB space is called the pixel
depth. Consider an RGB image in which each of the red, green, and blue images
is an 8-bit image. Under these conditions each RGB color pixels [that is,a triplet
of values (R,G,B)] is said to have a depth of 24-bits. (3 image planes times the
number of bits per plane). The term full color image is used often to denote a
24-bit RGB color image. The total number of colors in a 24-bit RGB image is
(28 )3 = 16, 777, 216 colors.
Red
Blue
The RGB color cube is a solid cube composed of the (28 )3 = 16, 777, 216
colors mentioned in the preceding paragraph. Figure 1.52 shows that an image
of the cross - sectional plane is viewed simply by feeding the three individual
component images into a color monitor. In the component images, 0 represents
black and 255 represents white (note that these are gray scale images).
It is of interest to note that acquiring a color image is basically the process
shown in Figure 1.52 in reverse. A color image can be acquired by using three
filters, sensitive to red, green and blue, respectively. When we view a color scene
with a monochrome camera equipped with one of these filters, the result is a
monochrome image whose intensity is proportional to the response of that filter.
Repeating this process with each filter produces three monochrome images that
are the RGB component images of the color scene. (In practice, RGB color image
sensors usually integrate this process into a single devices). Clearly, displaying
these three RGB component images in the form shown in Figure 1.52 would yield
an RGB color rendition of the original color scene.
While high end display cards and monitors provide a reasonable rendition of
the colors in a 24-bit RGB image, many systems in use today are limited to 256
colors. Also, there are numerous applications in which it simply makes no sense
to use more than a few hundred, and sometimes fewer, colors. A good example of
this is provided by the pseudocolor image processing techniques discussed. Given
the variety of systems in current use, it is of considerable interest to have a subset
of colors that are likely to be reproduced faithfully, reasonably independently of
viewer hardware capabilities. This subset of colors is called the set of safe RGB
colors, or the set of all - systems - safe colors. In internet applications, they are
called safe Web colors or safe browser colors.
On the assumption that 256 colors is the minimum number of colors that can
be reproduced faithfully by any system in which a desired result is likely to be
displayed, it is useful to have an accepted standard notation to refer to these color.
Forty of these 256 colors are known to be processed differently by various oper-
ating systems, leaving only 216 colors that are common to most systems. These
216 colors have become the defact standard for safe colors, especially in Internet
applications.
Number System Color Equivalents
Hex 00 33 66 99 CC FF
Decimal 0 51 102 153 204 255
Table 1.3 Valid values of each RGB component in a safe color
Each of the 216 safe colors is formed from three RGB values as before, but
each value can only be 0, 51,102,153,204 or 255. Thus RGB triplets of these
values give us (6)3 = 216 possible value (note that all values are divisible by 3).
Digital Image Fundamentals 1.55
Figure 1.53 shows the 216 safe colors, organised in a cube. The square in the
top left array has value FFFFFF (white), the second square to its right has value
FFFFCC, the third square has value FFFF99, and so on for the first row. The
second row of that same array has values FFCCFF, FFCCCC, FFCC99, and so
on. The final square of that array has value FF0000(the brightest possible red).
The second array to the right of the one just examined starts with value. CCFFFF
and proceeds in the same manner, as do the other remaining four arrays. The final
(bottom right) square of the last array has value 000000 (black). It is important to
note that not all possible 8 bit gray colors are included in the 216 safe colors.
Figure 1.53 shows the RGB safe - color cube. Unlike the full - color cube
which is solid, the cube in Figure 1.53 has valid colors only on the surface planes.
As shown in Figure 1.53, each plane has a total of 36 colors, so the entire surface
of the safe - color cube is covered by 216 different colors, as expected.
where again, the assumption is that all color values have been normalized to the
range[0,1]. Eq(1.23) demonstrates that light reflected from a surface coated with
pure cyan does not contain red (that is, C = 1−R in the equation). Similarly, pure
magenta does not reflect green, and pure yellow does not reflect blue. Equation
(1.23) also reveals that RGB values can be obtained easily from a set of CMY
values by subtracting the individual CMY values from 1. As indicated earlier, in
image processing this color model is used in connection with generating hard copy
output, so the inverse operation from CMY to RGB generally is of little practical
interest.
According to Figure 1.50, equal amounts of the pigments, cyan, magenta, and
yellow should produce black. In practice, combining these colors for printing
produces a muddy looking black. So, in order to produce true black (which is the
predominant color in printing) a fourth color, black, is added, giving rise to the
CMYK color model. Thus when publishers talk about “four - color printing,” they
are referring to the three colors of the CMY color model plus black.
As we have seen, creating colors in the RGB and CMY models and changing from
one model to the other is a straight forward process. As noted earlier, these color
systems are ideally suited for hardware implementations. In addition, the RGB
system matches nicely with the fact that the human eye is strongly perceptive to
red, green and blue primaries. Unfortunately, the RGB, CMY and other similar
color models are not well suited for describing colors in terms that are practical
for human interpretation. For example, one does not refer to the color of an au-
tomobile by giving the percentage of each of the primaries composing its color.
Further more, we do not think of color images as being composed of three primary
images that combine to form that single image.
When humans view a color object, we describe it by its hue, saturation, and
brightness. Recall from the discussion that hue is 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. It embodies the
achromatic notation of intensity and is one of the key factors in describing color
sensation. Thus HSI (Hue, Saturation, Intensity) color model an ideal tool for de-
veloping image processing algorithms based on color descriptions that are natural
and intuitive to humans.
Digital Image Fundamentals 1.57
White
I=0.75
Green
n Yel
Cya lo w
I=0.5 Blue I
Red
Blue Magenta
Black
Intensity
Intensity value can be easily extracted from the RGB image. This is done by
making the RGB cube stand on the black (0,0,0) vertex with the white (1,1,1)
vertex diversity above it, as shown in Figure 1.55. In this figure the line joining
the black and white vertices is called the intensity axis.
Thus, to determine the intensity component of any color, simply pass a plane
perpendicular to the intensity axis containing the color. The point at which the
plane intersects the intensity axis will give the intensity value in the range [0,1].
Saturation
From the Figure 1.55, it is clear that, as the planes move up and down the intensity
axis, the boundaries defined by the intersection of each plane with the faces of the
cube will have either a triangular or hexagonal shape, as shown in Figure 1.55. In
the hexagon, the primary colors are separated by 120 ◦ . The secondary colors are
60◦ from the primaries.
Consider an arbitrary color point inside the triangle shown as a dot. For this
color point, the saturation is the length of the vector from the origin to the color
point.
Hue
The Hue of a color point is determined as follows. The red axis is selected as the
reference axis. The angle ‘H’ made by the color point with the red axis represents
the Hue value.
1.58 Digital Image Processing
with
1
[(R − G) + (R − B)]
2
θ = cos−1 h i1/2 (1.25)
(R − G)2 + (R − B) (G − B)
1
I= (R + G + B) (1.27)
3
It is assumed that the RGB values have been normalized to the range [0,1] and
that angle θ is measured with respect to the red axis of the HSI space, as indicated
in Figure 1.54. Hue can be normalized to the range [0,1] by dividing the values
resulting from Eq(1.24) by 360◦ . The other two HSI components already are in
this range if the given RGB values are in the interval [0,1].
Problems
Problem 1.1
Let V = {0, 1} compute D4 , D8 and De distance between p & q.
y
1(q) 1 2 3
0 2 2 1
1 1 0 2
2 1 2 1(p)
x
Fig 1.56
Solution
The Euclidean distance ‘De ’ is calculated from the coordinates of p & q. the
coordinates of ‘p’ are (3,3) & the coordinates of ‘q’ are (0,0)
1. Euclidian distance ‘De ’ = (x − s)2 + (y − t)2
p
p
= (0 − 3)2 + (0 − 3)2
p
= (3)2 + (3)2
√
= 9+9
√
= 18
√
=3 2
2. D4 Distance
D4 = |(x − s)| + |(y − t)|
D4 = |(0 − 3)| + |(0 − 3)|
D4 = 6
3. D8 Distance
D8 = max [|x − s| , |y − t|]
D8 = max [|0 − 3| , |0 − 3|]
Digital Image Fundamentals 1.61
D8 = max(3, 3)
D8 = 3
Problem 1.2
Consider two image subsets S1 and S2
S1 S2
0 0 0 0 0 0 0 0 2 2
2 0 0 2 0 0 2 0 0 2
2 0 2 0 2 2 0 0 0
q
0 0 2 2 p 2 0 0 0 0 0
0 0 2 2 2 0 0 2 2 2
Fig 1.57
Problem 1.3
Find out D4 and D8 of 5 × 5 image given in Figure 1.58.
3 2 4 3 1q
0 4 4 3 2
2 2 2 0 2
2 2 1 1 1
p 1 0 1 0 3
Fig 1.58
1.62 Digital Image Processing
Solution
p(x1 , y1 ) = p(4, 0)
q(x2 , y2 ) = q(0, 4)
Problem 1.4
Consider the image segment shown below for V = 0, 1, compute the length of the
shortest m path between p and q.
3 1 2 1(q)
2 2 0 2
1 2 1 1
1(p) 0 1 2
Solution
The path taken from the image segment is equal to 5.
3 1 2 1
2 2 0 2
1 2 1 1
1 0 1 2
Fig 1.59
Digital Image Fundamentals 1.63
Problem 1.5
Consider an observer looking at a lamp post which is at a distance of 50m. If the
height of the lamp post is 10m find the size of the image in the retina, if the focal
length is 17mm.
10m
50m 17mm
Fig 1.60
Solution
10 r
=
50 17
10
r= × 17
50
r = 3.4mm
Problem 1.6
Consider the image subsets S1 and S2 given below for V = 4, determine whether
S1 and S2 are
a. 4 - Connected b. 8 - Connected c. m - Connected
S1 S2
2 2 2 2 2 2 2 4 4 2
4 2 2 4 2 2 4 2 2
4 2 2 4 2 4 4 2 2 2
2 2 4 4 4 2 2 2 2 2
2 2 4 4 4 2 2 4 4 4
Fig 1.61
Solution
Problem 1.7
A 4 × 4 sub image is shown in below Figure (1.62). Let V = 2, 4 and compute
the D4 , D8 and Dm distance between p and q.
y
4 2 2(p) 3
4 3 2 1
1 2 2 0
2(q) 3 1 0
x
Fig 1.62
• Image Acquisition
• Image enhancement and restoration
• Segmentation
• Color image processing
• Morphological processing
• Representation and Description
• Recognition and Interpretation
• Image compression.
6. What is morphological processing?
Morphological processing deals with tools for extracting image components
that are useful in the representation and description of shape.
7. Specify the elements of DIP system.
• Image Sensors
• Specialized image processing hardware
• Image processing software
• Computer
1.66 Digital Image Processing
• Mass storage
• Image displays
8. Explain the types of digital storage.
• Short term storage for use during processing , ex: computer memory,
frame buffers etc.,
• Online storage for relatively fast recall, ex: Magnetic disks or optical
media storage
• Archival storage for rare access, ex : Magnetic tapes, juke boxes.
9. What are the membranes that enclose the eye?
• Cones and
• Rods
11. Define brightness (or) subjective brightness.
Brightness is a subjective descriptor that is practically impossible to mea-
sure. Subjective brightness of an object is the luminance or intensity as
perceived by the Human Visual System(HVS).
12. What is brightness adaptation?
HVS (Human Visual System) cannot operate over a large range of bright-
ness simultaneously. Rather to accomplishes this large variation by changes
in its overall sensitivity, this phenomenon is known as brightness adapta-
tion.
13. Differentiate photopic and scotopic vision. (or) Differentiate cones and
rods.
Actual intensity
Perceived intensity
Fig 1.63
Fig 1.64
• RGB model— used for color monitor and color video camera
• CMY/CMYK model—used for color printing
• HSI model—-used for color image processing
• YIQ model—used for color picture transmission
Digital Image Fundamentals 1.69
31. Find the number of bits required to store a 256 × 256 image with 32
gray levels.
Given : 32 gray levels (i.e., L = 32)
L = 2k = 32, therefore k = 5 bits
256 ∗ 256 ∗ 5 = 3, 27, 680 bits.
32. What do you meant by Zooming and shrinking of digital images?
Zooming may be viewed as over sampling. It involves the creation of new
pixel locations and the assignment of gray levels to those new locations.
Shrinking may be viewed as under sampling. To shrink an image by one
half, we delete every alternate row and column.
33. Define N4 (p).
A pixel p at coordinates (x, y) has four horizontal and vertical neighbors
whose coordinates are given by (x + 1, y), (x − 1, y), (x, y + 1), (x, y − 1).
This set of pixels are called 4-neighbors of p denoted by N 4 (p).
34. Define ND (P ).
The four diagonal neighbors of p have coordinates (x + 1, y + 1),
(x + 1, y − 1), (x − 1, y + 1), (x − 1, y − 1) and are denoted by N D (p).
• 4 -adjacency
• 8-adjacency
• m-adjacency
Types of connectivity.
(a) 4 connectivity
(b) 8 connectivity
(c) m-connectivity (mixed connectivity)
38. What is meant by path or curve ?
A digital path or curve from pixel p to pixel q is a sequence of distinct
pixels.If p = q then the path is closed path.
1.72 Digital Image Processing
39. Give the formula for calculating euclidean distance, D4 (city block dis-
tance), D8 (chess board distance).
Consider two pixels p, q with co-ordinates (x, y) and (s, t) respectively,
three types of distances are defined as
Euclidean distance, De (p, q) = (x − s) + (y − t)
p
Fig 1.68
If the two dimensional function (i.e., the image) is under sampled then a
phenomenon called aliasing corrupts the sampled images. The corruption
is in the form of additional frequency components being introduced into
the sampling function. These are called aliasing frequencies. The effect of
aliased frequencies can be seen in the form of ‘Moire Patterns’ as shown in
Figure 1.65.
41. What is false contouring?
If the number of quantization levels is not sufficient a phenomenon called
false contouring becomes visible. When groups of neighboring pixels are
quantized to the same value, regions of constant gray levels are formed,
whose boundaries are called contours.
42. How to overcome the effect of false contouring?
False contouring can be held below the level of visibility by using the fol-
lowing two methods.
1. Contrast quantisation 2. Pseudorandom quantisation (Dither).
43. What is dither (pseudorandom quantisation)?
Dither is a method used to reduce the effect of false contouring. A small
amount of uniformly distributed pseudorandom noise is added to the lu-
Digital Image Fundamentals 1.73
Review Questions
1. Explain the steps involved in digital image processing. (OR) Explain vari-
ous functional blocks of digital image processing.
2. Describe the elements of visual perception and image formation in the eye.
3. With necessary diagrams explain sampling and quantization. (OR) with
necessary diagrams explain how an analog image is converted into digital
image.
4. Describe the components of digital image processing system with neat dia-
gram.
5. What are the elements of an image processing system and describe its work-
ing? How this is used for weather forecasting applications?
6. Explain the basic relationships between pixels?
7. What is a color model? What are its types? Explain RGB and HSI color
models with necessary diagrams.
8. Explain digital camera working principle.
9. With necessary diagram explain the construction and principle of operation
of a vidicon camera tube.
10. What is a frame buffer ? Discuss the categories of digital storage for image
processing applications.
11. What is meant by the image sensing? Explain in detail the construction and
operation of various image sensing and acquisition devices.