GODAVARI INSTITUTE OF

ENGINNERING AND
TECHNOLOGY

DEPARTMENT OF COMPUTER SCIENCE AND

ENGINEERING

A PAPER PRESENTION ON
IMAGE PROCESSING
BY

P.S.D ALEKHYA G.V.ALEKHYA

4/4CSE 4/4 CSE

alekhya.pinninti@gmail.com
 techniques that can
identify shades,
colors and
relationships that
cannot be perceived
by the human eye.
ABSTRACT
Image processing is 2. Any image

a technique to enhance improvement, such as

raw images received from refining a picture in a

cameras/sensors placed paint program that

on satellites, space probes has been scanned or

and aircrafts or pictures entered from a video

taken in normal day-to- source.

day life in various 3. A set of

applications. The data computational techniques

from an image is for analyzing, enhancing,

digitized and various compressing, and

mathematical operations reconstructing images

are applied to the data, 4. Used to solve

generally with a digital identification

computer, in order to problems, such as in

create an enhanced image forensic medicine or in

that is more useful or creating weather maps

pleasing to a human from satellite pictures. It

observer, or to perform deals with images in

some of the interpretation bitmapped graphics

recognition tasks usually format that have been

performed by humans.. scanned in or captured

with digital camera.
It is
Various
1. The analysis of a
techniques have been
picture using
developed in image  May remove
processing during the last noise,
 Improve the
four to five decades.
contrast of the
Most of the techniques image,
are developed for  Remove blurring
caused by
enhancing images movement of the
obtained from unmanned camera during
image acquisition
aircrafts, space probes It may correct for
and military geometric
reconnaissance flights. A few decades ago,
Image processing systems image processing was
are becoming popular due done largely in the analog
to easy availability of domain, chiefly by
powerful personal optical devices. Analog
computers, large size image processing refers
memory devices, graphic to the alteration of image
software’s etc through electrical means.
These optical methods
INTRODUCTION
Image processing is in are still essential to
many cases concerned applications such as
with taking one array of
pixels as input and holography because they
producing another array are inherently parallel;
of pixels as output, which
in some way represents however, due to the
an improvement to the significant increase in
original array. Most
image processing computer speed, these
techniques involve techniques are
treating the image as a
two-dimensional signal increasingly being
and applying standard replaced by digital image
signal processing
techniques to it. processing methods.
For example, this
processing Digital image processing
techniques are generally
more versatile, reliable,  Resolution
and accurate; they have
 Dynamic range
the additional benefit of
being easier to implement  Bandwidth
than their analog  Filtering
counterparts. Digital  Differential
computers are used to operators
process the image. The
 Edge detection
image will be converted
to digital form using a  Domain
digitizer and then process modulation.
it. Today, hardware
solutions are commonly Two-dimensional
used in video processing techniques
systems. However,  Image
commercial image representation
processing tasks are more  Image
commonly done by preprocessing
software running on
 Image
conventional personal
enhancement
computers.
 Image restoration
Commonly used signal  Image analysis
processing techniques:  Image
Most of the signal reconstruction
processing concepts that  Image data
apply to one-dimensional
compression
signals also extend to the
Image Reresentation :
two-dimensional image
image defined in the to
signal. Some of these
be a function of two real
one-dimensional signal
variables, for example, f
processing concepts
(x, y) with f as he
become significantly
amplitude (e.g.
more complicated in two-
brightness) of the image
dimensional processing.
at the real coordinate
Image
position (x, y).
processing brings some
new concepts, such as
connectivity and
rotational invariance that
are meaningful only for
two-dimensional signals.

One-dimensional
techniques
 Image resolution can be
The 2D continuous measured in various
image f (x, y) is divided ways. Basically,
into N rows and M resolution quantifies how
columns. The intersection close lines can be to each
of a row and a column is other and still be visibly
called s pixel. The value resolved. Resolution units
assigned to the integer can be tied to physical
coordinates [m, n] with sizes (e.g. lines per mm,
{m=0,1,2,.........M-1} and lines per inch) or to the
{n=0,1,2,.............N-1} is overall size of a picture
f {m, n}. In fact in most (lines per picture height,
cases f (x, y) -- we might also known simply as
consider to be the lines, or TV lines).
physical signal that Furthermore, line pairs
impinges on the face of are often used instead of
the sensor. Typically an lines. A line pair is a pair
image file such as BMP, of adjacent dark and light
JPEG, TIFF etc., has lines, while lines count
some picture and header both dark lines and light
information. A header lines. A resolution of 10
usually includes details lines per mm means 5
like format identifier dark lines alternating
(typically first with 5 light lines, or 5
information), resolution, line pairs per mm.
number of bits/pixel, Photographic lens and
compression type etc. film resolution are most
often quoted in line pairs
per mm.
Image resolution
Image resolution
describes the detail an Scaling:
image holds. The term
applies equally to digital
images, film images, and
other types of images.
Higher resolution means
more image detail. It is
illustrated by following
image.
Scaling is defined as the row, mth column of the
increase or original imagery is
reduction of image selected and displayed.
size by a fixed Another way of
ratio. The theme of accomplishing the same
the technique of is by taking the average
magnification is to in ‘m x m’ block and
have a closer view displaying this average
by magnifying or after proper rounding of
zooming the the resultant value.
interested part in
the imagery. By
reduction, we can
bring the
unmanageable size
of data to a
manageable limit.
Image
Preprocessing

1. Magnification
This is usually done to
improve the scale of
display for visual 2-dimensional image
interpretation or scaling
sometimes to match the
scale of one image to In the above signal, the
other. To magnify an image is magnified
image by a factor 2, each vertically and reduced
pixel of the original horizontally.
image is replaced by Rotation
block of 2 x 2 pixels, all Image rotation is
with the same brightness performed by computing
value as the original the inverse
pixel. transformation for every
destination pixel.
2. Reduction Rotation is used in image
Image reduction increases mosaic, image
the incidence of high registration etc. One of
frequencies and causes the techniques of rotation
several pixels to collapse is 3-pass shear rotation,
into one. To reduce a where rotation matrix can
digital image of the be decomposed into three
original data, every mth separable matrices.
 Image rotation

Image mosaicking
Image enhancement:
Mosaic Image enhancement is
the improvement of
digital image quality,
without knowledge about
the source of degradation.
In Image enhancement,
the goal is to accentuate
certain image features for
subsequent analysis or for
image display. If the
source of degradation is
known, one calls the
process image
restoration. Both are
iconical processes, viz.
input and output are
images. It is quite
easy, for example, to
make an image lighter or
darker, or to increase or
decrease contrast, pseudo
colouring, noise filtering,
sharpening and
magnifying. Advanced Image
image enhancement
software also supports
enhancement
many filters for altering
images in various ways. An image is enhanced
Programs specialized for when it is modified so
image enhancements are that the information it
sometimes called image contains is more clearly
editors. evident, but enhancement
can also include making
The aim of image the image more visually
enhancement is to appealing. An example is
improve the noise smoothing.
interpretability or Another example of
perception of information enhancement is contrast
in images for human manipulation, where each
viewers or to provide pixel's value in the new
`better' input for other image depends solely on
automated image that pixel's value in the
processing techniques. It old image; in other
is illustrated as follows: words, this is a point
operation. Contrast
manipulation is
commonly performed by
adjusting the brightness
and contrast
Image noise reduction:
Images taken with both
digital cameras and
conventional film
cameras will pick up
noise from a variety of
(partially) removed - for
aesthetic purposes as in compensate for the
artistic work or degradation it caused.
marketing, or for The most common
practical purposes such as degradations have their
computer vision origin in imperfections of
the sensors, or in
Noise filtering is used to transmission.
filter the unnecessary
information from an
image. It is also used to
remove various types of
noises from the images.

Noise reduction

Image restoration:
Images typically suffer
from a range of
imperfections including
geometric distortion, non-
uniform contrast, and
noise. Image restoration Image restoration
removes or minimizes
some known degradations The aim of restoration is
in an image. It can be also to improve the
seen as a special kind of image, but unlike
image enhancement. enhancement, knowledge
Image restoration is to of how the image was
"compensate for" or formed is used in an
"undo" defects, which attempt to retrieve the
degrade an image. ideal (uncorrupted)
Degradation comes in image. Any image-
many forms such as forming system is not
motion blur, noise, and perfect, and will
camera misfocus. In cases introduce artifacts (for
like motion blur, it is example, blurring,
possible to come up with aberrations) into the final
a very good estimate of image that would not be
the actual blurring present in an ideal image.
function and "undo" the A point-spread function,
blur to restore the original called a filter, can be
image. In cases where the constructed that undoes
image is corrupted by the blurring. By imaging
noise, the best we may the blurred image with
hope to do is to the filter point spread
function, the restored
image results. The filter
point spread function is systems measure
spread out more than the quantitative information
blurring point spread and use it to make a
function, bringing more sophisticated decision,
pixels into the averaging such as controlling the
process. This is an arm of a robot to move an
example of a global object after identifying i
operation, since perhaps or navigating an aircraft
all of the pixels of the with he aid of images
blurred image can acquired along its
contribute to the value of trajectory.
a single pixel in the
restored image. This type Image Reconstruction:
of deblurring is called Image reconstruction is a
inverse filtering, and is special class of image
sensitive to the presence restoration, where a two-
of noise in the blurred dimensional object is
image. By modifying the reconstructed from
deblurring filter several One-dimensional
according to the projections. Each
properties of the noise, projection is obtained by
performance can be projecting a parallel X-
improved. An example of ray (or other penetrating
the need to deblur images radiation) beam through
from an optical system is the object. Planar
the Hubble Space projections are thus
Telescope before its obtained by viewing the
spherical aberration was object from many
corrected with new different angles.
optics. Reconstruction
Image Analysis: algorithms derive an
Image analysis is image of a thin axial slice
concerned with making of the object, giving an
quantitative inside view otherwise
measurements from an unobtainable without
image to produce a performing extensive
description of it. In the surgery. Such techniques
simplest form, this ask are important in medical
could be reading a label imaging (ct scanners)
on a grocery item, sorting astronomy, radar
different pars on an imaging, geological
assembly line, or exploration, and non-
measuring he size ad destructive testing of
orientation of blood cells assemblies.
in medical image. More
advanced image analysis Image Compression:
Image Compression as three to four may be
(often known as Image acceptable.
Coding) is the art /
science of representing
images with the least
information (no. of bits)
consistent with achieving
an acceptable image
quality / usefulness.
Compression is a way of
representing an image by
fewer numbers, at the
same time minimizing the
degradation of the
information contained in
the image. Compression
is important because of
the large quantities of
digital imagery that are
sent electronically and
stored. Digital high-
definition television relies
heavily on image
compression to enable Image compression
transmission and display
of large-format color Image compression can
images. Once the image be lossy or lossless.
is compressed for storage Lossless compression is
or transmission, it must sometimes preferred for
be uncompressed for use, artificial images such as
by the inverse of the technical drawings, icons
compression operations. or comics. This is
There is a trade-off because lossy
between the amount of compression methods,
compression and the especially when used at
quality of the low bit rates, introduce
uncompressed image. compression artifacts.
High compression rates Lossless compression
are acceptable with methods may also be
television images, for preferred for high value
example. However, content, such as medical
where high image quality imagery or image scans
must be preserved (as in made for archival
diagnostic medical purposes. Lossy methods
images), only are especially suitable for
compression rates as low natural images such as
photos in applications term that covers the use
where minor (sometimes of digital image
imperceptible) loss of processing techniques to
fidelity is acceptable to process, analyze and
achieve a substantial present images obtained
reduction in bit rate. from a microscope.
The best image quality at
a given bit-rate (or Medical imaging:
compression rate) is the Medical imaging
main goal of image designates the ensemble
compression. of techniques and
processes used to create
APPLICATIONS: images of the human
 Photography and body for clinical purposes
printing or medical science.
 Satellite image
Morphing:
processing
Morphing is a special
 Medical image affect in motion pictures
processing and animations that
 Face detection, changes (or morphs) one
feature detection, image into another
face identification through a seamless
 Microscope image transition. Most often it is
processing used to depict one person
turning into another
 Car barrier
through some magical or
detection
technological means.
 Morphological
image processing

Face detection:
Face detection is a
computer technology that
determines the locations
and sizes of human faces
in arbitrary (digital) Morphing
images. It detects facial
features and ignores
anything else, such as
buildings, trees and
bodies. RESEARCH:
Image processing is an
Microscope image active area of research in
processing: such diverse fields as
Microscope image medicine, astronomy,
processing is a broad
microscopy, seismology, turnaround, easing of
defense, industrial quality growing workload, etc...
control, and the By embracing the new
publication and image processing
entertainment industries. technologies and further
The concept of an image refinements in image
has expanded to include processing techniques,
three-dimensional data users are likely to find it
sets (volume images), and more beneficial, not less,
even four-dimensional in future, while more
volume-time data sets. refinements in image
An example of the latter processing techniques
is a volume image of a will be appreciated at a
beating heart, obtainable reduced cost. Overall,
with x-ray computed Image processing is a
tomography (CT), good option that deserves
ultrasound, confocal a careful look.
microscopy, scanning
tunneling microscopy, RESULT:
atomic force microscopy,
and other modalities have In recent years there were
been developed to many improvements in
provide digitized images the areas of image
directly. Digital images processing, more
are widely available from specifically, the areas of
the Internet, CD-ROMs, block artifact and
and inexpensive charge- mosquito noise
coupled-device (CCD) reduction... Other typical
cameras, scanners, and areas of image processing
frame grabbers. Software improvements include
for manipulating images adaptive contrast
is also widely available. enhancement, sharpness
and texture enhancement,
CONCLUSION: and selective color
We all know that new correction. The latest
technologies, at first advancements also
perceived as unwelcome, resulted in the
subjected to some compression rate in
modifications, usually recent years there were
lead to better, more many improvements in
productive situations for the areas of image
everyone. In the case of processing, more
image processing, the specifically, the areas of
potential benefits are block artefact and
many: refined images can mosquito noise
be obtained, faster report reduction... Other typical
areas of image processing
improvements include
adaptive contrast
enhancement, sharpness
and texture enhancement,
and selective colour
correction. The latest
advancements also References: -
resulted in the Digital Image
compression rates, which Processing by Rafael
C.Gonzalez and Richard
turned up to squeeze as
E.Woods.
much video content as
Hoose, N. : ‘Computer
possible in to available
Image Processing in
distribution channels and
Traffic Engineering’.
onto media.
Rourke, A., and Bell,
Advanced compression
M.G.H.: ‘Queue detection
techniques are replacing and congestion monito
previous-generation ring using
technology, offering mage processing’, Traffic
enhancements such as Engg. and Control.
better streaming Traffic Queue Length
capability, higher Measurement Using an
compression for a given Image Processing Sensor
quality and lower latency. by Masakatsu
As a result of all these Higashikobo, Toshio
techniques image Hinenoya and Kouhei
processing is including Takeouchi.
motion compensated A Real-time Computer
processing in itself. As it Vision System for Vehicle
is incorporating new Tracking and Traffic
levels of processing Surveillance by
(based on motion Benjamin Coifman
compensation) we will be (corresponding author).
able to see a wide variety A Real-time Computer
of offerings and various Vision System for
levels of quality and Measuring Traffic
capabilities. Parameters by David
Beymer, Philip
McLauchlan, Benn
Coifman and Jitendra Ma