INTRODUCTION TO IMAGE PROCESSING 23

1.8.1 Surveyof Image Processing Applications Based on EM Radiation

The survey from the electromagnetic domain is justified as visible light is part of the
electromagnetic spectrum. Electromagnetic waves are waves of energy that have both
electric and magnetic characteristics and which can travel at the speed of light in vacuum.
Electromagnetic waves can be classified based on the frequency or wavelength and its
select parts are shown in Table 1.2,
Table 1,2 Select parts of the electromagnetic spectrum

Types of radiation Frequency range Wave length Nature of imaging and its relevance
(in Hertz) (in cm) for image processing
Radio waves 10$-100 > 10 AM/FM radio

Microwave 101°-10!2 10°-106 Radar imaging

Infrared 102-1014 10°-7000 Thermal imaging
Visible light 4-7.5 x 1014 7000 4000 Optical
Ulraviolet 10'5-10!7 4000-10 Optical
101"-1020 10-0.1 Medical and industrial
X-rays
1020-1024 <0.1 Medical
Gamma rays
Image processing applications fully exploit the entire range of the electromagnetic
spectrum and some of the applications are given here.
Radio waves Radio waves have the lowest energy level and are used in radio and video
broadcast as well as by mobile phones. Radio waves are used extensively in radios, satellites,
radar, and computer networks for transmission of data. In image processing domain, radio
waves are useful in two areas--remnote sensing and medical imaging. Hydrogen gas in
radar systems,
Space releases energy in low frequency and is collected as radio waves. In
and is
the released energy is collected back, which helps to image the surface of the earth
used in the
useful in predicting weather forecasts. In medical imaging, radio waves are
Tom of magnetic resonance imaging (MRI). Apatient is placed on a powerful magnet, and
responding
hen radio waves are passed through the patient's body in short pulses. The location
determines the
pulses of the radio waves are collected by the computer, which
and image
Of these signals and their strength. MRI is capable of detecting soft tissues,
processing applications can be used for analysing MRI images.
an abbreviation
MIcrowaves Microwaves are useful in radar and medical imaging. Radar is
microwaves in addition to radio
O KAdio Detection And Ranging. Radar systems use
forecasting, imaging
waves at higher frequency and are used in applications such as weather
is a
Planetary surfaces, and determining earth resources. Synthetic aperture radar (SAR)
microwaves is used in medical
tor detecting targets and its velocity. The concept of electrons. This phenomenon
iging toprobe the magnetic resonance (MR) properties of
 DIGITAL IMAGE PROCESSING
24

resonance (ESR) and electron paramagnetic resonance (EPR), and is

is called elcctron spin biological images.
used for measurenment of free radicals in
range
Infrared waves The rangeof infrared
(IR) radiation is about Imm-750 nm. The
the visible spectrum is called 'ncar infrared' and the larger wavelength is called
adjacent to EM radiation and the quantity.
*far infrared'. All objects at afixed temperature can emit
The domain of thermal
type, and distribution of wavelength depend on temperature.
body surfaces. Night
imaging involves the use of temperature images of the human
vision technology is another major domain of infrared waves. In night vision, one can see
objects in the absence of light. The military, world over, uses infrared cameras, to detect
objects at night. Another domain is thernmographic imaging, where the images are called
themograms. Infrared cameras can penetrate smokes to detect locations, moving objects,
and fleeing criminals. Thermograms are also used in military reconnaissanceapplications.
In medical imaging, infrared is useful for applications such as detecting breast cancer and
in locating hot and cold regions of the human body.
Visible light region Visible light is part of the EM spectrum that the human eye can
perceive and it varies between 750 nm and 400nm. Most of the traditional image processing
applications such as photography, image archiving, and content-based retrieval systemsare
in the visible light range only. Telescopes are also examples of how
for astronomy applications. The medical domain imaging can be used
are endoscopy and dermoscopy. Endoscopy is used for
applications in the visible light range
Microscopy images that are used to study the strain tissueexamining
the inner body parts.
range.
patterns also fall in the visible
Ultraviolet rays Ultraviolet rays are used in applications such as
lasers, biomedical imaging, and lithography, microscopy,
astronomical observations.
X-rays X-rays are
high-frequency
They are used in a number of medical electromagnetic rays.
X-ray imaging helps doctors in imaging applications.
effective diagnosis. Figure 1.16 providing vital details for
Fig. 1.16 X- ray image image. Some of the other imageillustrates chest X-ray
a
are analysis of processing applications
electrophoresis
bacterial colony counting, and gels, chromosome analysis,
smallest wavelengths. autoradiography
Gamma rays Gamma rays have the analysis.
energy in comparison with other They also have the highest
gamma rays which are waves in the EM
spectrum. Radioactive elements emit
with an isotope, whichused extensively in nuclear
emits gamma rays as it medicine. Normally, a patient is injected
the gamma ray
detector and an image is formed.decays. These
This kind of emissions are collected by
imaging. Gamma rays are also used in
for finding tinyimaging in called invasive
is
and in astronomical
observations. industries cracks metal surfaces
 INTRODUCTION TO IMAGE PROCESSING 25

Apart from EM, sound waves can be used by imaging devices. One example is ultrasound
imag1ng.
Ultrasound imaging Sound is a longitudinal wave and it is possible to detect objects and
produce images with sound.

GB

Fig.1.17 Ultrasound image

Like radar applications, a transducer can send sound waves at shorter burst. The
reflected sound waves are picked by sensors and the time interval between the sending
and receiving of waves can be used to locate the objects. Figure 1.17 shows an ultrasound
image of a gallstone formed in the gallbladder, acommon disorder affecting the biliary
system. Ultrasound images have very good temporal resolution, good tissue contrast and
its low cost makes it useful for many medical imaging applications.
Another important application using sound waves is SONAR. SONAR is an abbreviation
of SOund Navigation And Ranging. It can be used to find underwater objects. Synthetic
aperture sonar can be used to locate the seabed and wrecks. It can be used to locate broken
objects of naval vessels and submarines.

182 Survey of lmage Processing Applications Based on Application Domain

Imaging applications can also be surveyed based on the domains where images are used.
The survey of applications involving image processing based on the applications domains
are given here.
Pattern recognition and biometrics Biometrics literally means measurement of life. This
Involves measuring and analysing physical attributes such as fingerprint, facial features,
DNA, and iris and behavioural attributes such as gait, voice, and signatures for verification
26 DIGITAL IMAGE PROCESSING

image-based
and identification purposes. With the concept of pattern recognition, many
of pattern
applications like object recognition is possible. We will discuss the conceptattendance
recognition in detail in Chapter 13. Other applications include biometrics-bascd
system, biometrics-based authentication system and biometric matching.
Medical imaging Medical imaging is onc arca
where image processing is fully utilizcd. Image
processing is very useful in interpreting medical
images, from simple diagnosis to advanccd
telesurgical applications. Telesurgical applications
are those where robots are used to perform surgical
operations. Asample foetal MRI image is shown in
Fig. 1.18.
Fig. 1.18 Foetal MRI image
Image visualization and rendering Image
visualization is agrowing area in image processing.
Image processing algorithms can assist in
visualization of images by converting a set of
2D images to form 3D images. Image rendering
algorithms on the other hand can render 3D images
to form 2D images. Figure 1.19 shows a CT Scan of
the skull.
The figure clearly shows anatomical structures
Fig. 1.19 CT skull image such as soft tissues and bones, which aid in diagnosis
and treatment. In general, image
visualization is
helpful in
medical students, planning of medical treatment,various applications such as in educating
and in telesurgery applications.
Industrial automation Machine vision systems based on
increased quality using automatic inspection and process image processing guarantee
inspection is a vast field where image processing is used bymonitoring. Automated visual
food, textiles, and plastic for
automated surface
industries such as aerospace,
mistakes because of the repeated and routine testing. As humans are bound to make
deployed by industries for applications such asnature
visual
of the job, many
automated systems are
measurement. inspection, surface defect testing, and
fiber analysis, Applications such as measurement of belt
and elimination of surface width, surface quality inspection,
factory automation. defects in aluminum foils are some examples of
Remote sensing The role of image
immense. Imaging applications in processing in remote sensing applications is qute
and prediction of
atmospheric meteorological
changes; domain include weather
locating natural resources such as forecasting
forests, vegetation areas, land cover, and terrain drinking water,
rendering: and automatic event detection.
 INTRODUCTION TO IMAGE PROCESSING 27

Many environmental monitoring applications have been devcloped to monitor large or

remote tractsof land such as desertsand forests. Cartogruphy applications also usc image
processing techniques to fix the boundaries of aregion. Aerial archeology is another
grow ing area where applications are being developed for imaging carth, weather forecasts
and environmental science wherce changes thut happen over time (such as agriculture, urban
developnnent, deserts) can be recorded and processed.
Image communication Image communication as a domain started with the transmission of
images by newspaper industry using the submarine cable between London and New York.
This ledto the needand developnent of inage enhancement and compression algorithms.
The developments in the domain of image compression made possible many image-bascd
communication applications such as tclevision and video broadcasting and transmission
of CCTV footages. The availability of low-cost cameras and thc Internet led to many
applications such as facsimile image transmission, video conferencing, and video phones.
is
Image security and copyright protection Media security, or multimedia sccurity,
a domain involving security of images, video, audio, and graphics. Many image-based
applications are developed involving the image sccurity concepts. Digital steganography
another image such
is a technique for data hiding where an image is conccaled within
domain that is of interest is
that no one can detect the presence of the image. Another
or more items of
digital watermarking which is similar to steganography in which onc
tamper the
information known as watermarks are added to the image. Any attempt to
technology can be useful for
images affects watermarks. Therefore digital watermarking
related technique, called
copyright identification, monitoring, and copyright protection. A content on the part of
change in the
digital signatures, is used to assert the ownership. Any
such example is shown in Fig. 1.20.
an unauthorized user invalidates the signature. One information as the illegal
Fingerprinting is another technique that asserts the purchaser
copies betray the purchaser.

(a)
Fig. 1.20 Watermarking (a) Original image
 DIGITAL IMAGE PROCESSING
28

Visible watermark demo

(b)
Fig. 1.20 (b) Image with watermark-Visible Watermark Demo'

Digital scrambling technique uses encryption algorithms such as substitution,

transposition and modular functions to scramble the image contents, and the original image
can be recovered using decryption algorithms. Digital scrambling algorithms are used
effectively by television industry. Virtualcryptography is another emerging area where the
user needs no knowledge of the cryptographic concepts and can be used to share secrets
using the concepts of image processing.
Image processing also helps in the authentication process using the concept of artificial
intelligence. CAPTCHA is an example of how image-based security applications can be
developed using the assumption that slanting letters, broken, and highly noisy
can make recognition by machines very background
complicated. However, at the same time, most
humans can solve this problem easily. This fact helps in the
for protecting websites and to prevent development of applications
automatic indexing of websites by search engines.
Digital image forensics Image forensics is an
authenticity of the given image. In many countries, emerging field that aims to validate the
as valid evidence. court of law permits images and video
However, with the proliferation of
manipulate images, tampering can be done with the applications which can be used to
intention. If tampering is done with malicious images with or without any malici0US
find whether the image is
authentic or not.
intention, then digital forensics can help to
Video processing Avideo signal is a
represent asequence of images. Thus,one-dimensional
onc can visualizetime-varying signal whose contenb
of image video
processing. Vidco processing applications processing as an extensiO
restoration, video compression, and video include video enhancement, video
Image understanding Image tracking.
inunderstanding
3D information from a or computer vision is
scenc
human eyes observe objects of the Stercoscopic imaging is one useful in recoverin?
2D.
scene and reconstruct to get simple example of no
the depth
informato
 INTRODUCTION TO IMAGE PROCESSING 29

They then use these reconstructed images to perceive the 3D world. Thc movie industry
uses this concept toproduce many 3D movies. Robotic vision aims at developing robotic
applications such as hurdle detection and bypassing.
Document image processing Document image proccssing aims to create a paperless office
by capturing the documents in the forn of images. Image archival systems are uscd to store
the document images which can be retrieved by the retrieval systems. This leads to some
interesting applications such ascontent-based retricval systems, image scarch engines, and
script recognition systems.
Image processing in military applications Many military rcconnaissance systems usc
image processing technology. Themal imaging systems have the ability to acquire useful
images at night and under adverse atmospheric conditions such as fog and smoke. This
heips military personnel involved in target acquisition and reconnaissance operations.
Computational photography and photography Photography is an excellent example of how
image processing is helpful to the common man. Image processing is helpful in creating
special effects such as warping, blending, animation, and other visual effects. Similarly,
publishers can enhance the layout of the pages by using digital images for publication.
High dynamic range imaging is one of the recent areas of research where many images
are combined to create images which have dynamic range such that the images look more
realistic.

Image and video analytics Image and video data can be analysed using data mining
algorithms for extracting knowledge. As images and videos have enormous amount of
data, data analytics can be used to extract quality knowledge that can be used to improve
can
the performance of the existing systems substantially. Thus, image and video analytics
enable development of knowledge-based systems and intelligent systems.
effect for
Visual effects Sometimes images can be manipulated to create a visual
by mapping
Cntertainment purposes. For example, polar transformation can be performed
system. One such polar
the transformation of image from the Cartesian to polar coordinate
transformation is shown in Fig. 1.21.

Fig. 1.21 Polar transformation (a) Original image

(b) Polar transform
 DIGITAL IMAGE PROCESSING
30

used for personal entertainment as well as for commercial

Such visual effects can be more images as
applications. Image compositioning is an operation that takes two or
an integrated result of all images as output. Digital imaging makes this
input and produces be used in web pages and imaging
casier and is useful in creating photographs that can pieces. These
into many
applications. This technique is achieved by dividing the image called
piecesare then combined in different combinations to create a single output image
an element.
by a
Image mosaicking is an operation where every group of pixels is approximated
given image. Acollection of smallimages gives an illusion of a larger image when viewed
from a distance. The small images called tile images can be placed either manually or
automatically. One such mosaic is shown in Fig. 1.22.

E
(a) (b)
Fig. 1.22 Image mosaic (a) Tile mosaic (b) Photo
mosaic
Entertainment Video editing, animation, and image
applicationsthat can be used for both personal as morphing are few of the many
well as commercial purposes.
Cinematography is one of the popular ones where the sequence of images is shown in
rapid successionto create movie effect. Both
the movie and television
applications such as visual effects and industry use image
background for name slides.
Image-based retrieval systems In text-based retrieval
involving search terms. Using search terms as key systems, the input is a query
Similarly, content-based image retrieval systems words, the systems retrieve the
to retrieve identical related images. Picture
or (CBIR) involve text and image content.
queries
capture, store, and retrieve medical images archival systems are used by hospitals to
that are relevaht for the
patients.