International Journal of Trend in Scientific Research and Development (IJTSRD)

Volume 6 Issue 5, July-August 2022 Available Online: www.ijtsrd.com e-ISSN: 2456 – 6470

Image Processing in the Current Scenario

Supriya Kumari
Assistant Professor, Branch-Computer Science & Engineering, Purnea College of Engineering, Purnia, Bihar, India

ABSTRACT How to cite this paper: Supriya Kumari

Image processing is the process of transforming an image into a "Image Processing in the Current
digital form and performing certain operations to get some useful Scenario" Published
information from it. The image processing system usually treats all in International
images as 2D signals when applying certain predetermined signal Journal of Trend in
Scientific Research
processing methods.
and Development
There are five main types of image processing: (ijtsrd), ISSN: 2456-
Visualization - Find objects that are not visible in the image 6470, Volume-6 | IJTSRD51728
Recognition - Distinguish or detect objects in the image Issue-5, August
Sharpening and restoration - Create an enhanced image from the 2022, pp.1682-1691, URL:
www.ijtsrd.com/papers/ijtsrd51728.pdf
original image
Pattern recognition - Measure the various patterns around the objects Copyright © 2022 by author (s) and
in the image International Journal of Trend in
Retrieval - Browse and search images from a large database of digital Scientific Research and Development
images that are similar to the original image Journal. This is an
Open Access article
KEYWORDS: image, transforming, processing, signals, objects, distributed under the
databases, operations, detect, retrieval terms of the Creative Commons
Attribution License (CC BY 4.0)
(http://creativecommons.org/licenses/by/4.0)
INTRODUCTION
Fundamental Image Processing Steps
Image Acquisition Compression
Image acquisition is the first step in image Compression is a process used to reduce the storage
processing. This step is also known as preprocessing required to save an image or the bandwidth required
in image processing. It involves retrieving the image to transmit it. This is done particularly when the
from a source, usually a hardware-based source. image is for use on the Internet.
Image Enhancement Morphological Processing
Image enhancement is the process of bringing out and Morphological processing is a set of processing
highlighting certain features of interest in an image operations for morphing images based on their
that has been obscured. This can involve changing the shapes.
brightness, contrast, etc.
Segmentation
Image Restoration Segmentation is one of the most difficult steps of
Image restoration is the process of improving the image processing. It involves partitioning an image
appearance of an image. However, unlike image into its constituent parts or objects. [3,4]
enhancement, image restoration is done using certain
mathematical or probabilistic models.[1,2] Representation and Description
After an image is segmented into regions in the
Color Image Processing segmentation process, each region is represented and
Color image processing includes a number of color
described in a form suitable for further computer
modeling techniques in a digital domain. This step
processing. Representation deals with the image’s
has gained prominence due to the significant use of
characteristics and regional properties. Description
digital images over the internet. deals with extracting quantitative information that
Wavelets and Multiresolution Processing helps differentiate one class of objects from the other.
Wavelets are used to represent images in various
degrees of resolution. The images are subdivided into Recognition
wavelets or smaller regions for data compression and Recognition assigns a label to an object based on its
for pyramidal representation. description.

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1682
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
Applications of Image Processing It helps to improve images for human interpretation
Medical Image Retrieval Information can be processed and extracted from
Image processing has been extensively used in images for machine interpretation
medical research and has enabled more efficient and
accurate treatment plans. For example, it can be used The pixels in the image can be manipulated to any
for the early detection of breast cancer using a desired density and contrast
sophisticated nodule detection algorithm in breast Images can be stored and retrieved easily
scans. Since medical usage calls for highly trained
image processors, these applications require It allows for easy electronic transmission of images to
significant implementation and evaluation before they third-party providers[9,10]
can be accepted for use. Discussion
Traffic Sensing Technologies Image processing is a method to perform some
In the case of traffic sensors, we use a video image operations on an image, in order to get an enhanced
processing system or VIPS. This consists of a) an image or to extract some useful information from it. It
image capturing system b) a telecommunication is a type of signal processing in which input is an
system and c) an image processing system. When image and output may be image or
capturing video, a VIPS has several detection zones characteristics/features associated with that image.
which output an “on” signal whenever a vehicle Nowadays, image processing is among rapidly
enters the zone, and then output an “off” signal growing technologies. It forms core research area
whenever the vehicle exits the detection zone. These within engineering and computer science disciplines
detection zones can be set up for multiple lanes and too.
can be used to sense the traffic in a particular Image processing basically includes the following
station.[5,6] three steps:
Besides this, it can auto record the license plate of the Importing the image via image acquisition tools;
vehicle, distinguish the type of vehicle, monitor the Analysing and manipulating the image;
speed of the driver on the highway and lots more.
Output in which result can be altered image or
Image Reconstruction report that is based on image analysis.
Image processing can be used to recover and fill in
the missing or corrupt parts of an image. This There are two types of methods used for image
involves using image processing systems that have processing namely, analogue and digital image
been trained extensively with existing photo datasets processing. Analogue image processing can be used
to create newer versions of old and damaged photos. for the hard copies like printouts and photographs.
Image analysts use various fundamentals of
Face Detection interpretation while using these visual techniques.
One of the most common applications of image Digital image processing techniques help in
processing that we use today is face detection. It manipulation of the digital images by using
follows deep learning algorithms where the machine computers. The three general phases that all types of
is first trained with the specific features of human data have to undergo while using digital technique are
faces, such as the shape of the face, the distance pre-processing, enhancement, and display,
between the eyes, etc. After teaching the machine information extraction.
these human face features, it will start to accept all
objects in an image that resemble a human face. Face In this lecture we will talk about a few fundamental
detection is a vital tool used in security, biometrics definitions such as image, digital image, and digital
and even filters available on most social media apps image processing. Different sources of digital images
these days.[7,8] will be discussed and examples for each source will
be provided. The continuum from image processing
Benefits of Image Processing to computer vision will be covered in this. Finally we
The implementation of image processing techniques will talk about image acquisition and different types
has had a massive impact on many tech organizations. of image sensors.[11,12]
Here are some of the most useful benefits of image
processing, regardless of the field of operation: Many of the techniques of digital image processing,
or digital picture processing as it often was called,
The digital image can be made available in any were developed in the 1960s, at Bell Laboratories, the
desired format (improved image, X-Ray, photo Jet Propulsion Laboratory, Massachusetts Institute of
negative, etc) Technology, University of Maryland, and a few other

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1683
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
research facilities, with application to satellite time, for some dedicated problems such as television
imagery, wire-photo standards conversion, medical standards conversion. As general-purpose computers
imaging, videophone, character recognition, and became faster, they started to take over the role of
photograph enhancement. The purpose of early image dedicated hardware for all but the most specialized
processing was to improve the quality of the image. It and computer-intensive operations. With the fast
was aimed for human beings to improve the visual computers and signal processors available in the
effect of people. In image processing, the input is a 2000s, digital image processing has become the most
low-quality image, and the output is an image with common form of image processing, and is generally
improved quality. Common image processing include used because it is not only the most versatile method,
image enhancement, restoration, encoding, and but also the cheapest.
compression. The first successful application was the
In 1972, the engineer from British company EMI
American Jet Propulsion Laboratory (JPL). They used
Housfield invented the X-ray computed tomography
image processing techniques such as geometric
device for head diagnosis, which is what is usually
correction, gradation transformation, noise removal,
called CT (computer tomography). The CT nucleus
etc. on the thousands of lunar photos sent back by the
method is based on the projection of the human head
Space Detector Ranger 7 in 1964, taking into account
section and is processed by computer to reconstruct
the position of the sun and the environment of the
the cross-sectional image, which is called image
moon. The impact of the successful mapping of the
reconstruction. In 1975, EMI successfully developed
moon's surface map by the computer has been a huge
a CT device for the whole body, which obtained a
success. Later, more complex image processing was clear tomographic image of various parts of the
performed on the nearly 100,000 photos sent back by
human body. In 1979, this diagnostic technique won
the spacecraft, so that the topographic map, color map
the Nobel Prize. Digital image processing technology
and panoramic mosaic of the moon were obtained,
for medical applications was inducted into the Space
which achieved extraordinary results and laid a solid
Foundation Space Technology Hall of Fame in 1994
foundation for human landing on the moon.
Digital filters are used to blur and sharpen digital
The cost of processing was fairly high, however, with
images. Filtering can be performed by:
the computing equipment of that era. That changed in
convolution with specifically designed kernels
the 1970s, when digital image processing proliferated
(filter array) in the spatial domain[25]
as cheaper computers and dedicated hardware became
masking specific frequency regions in the
available. This led to images being processed in real-
frequency (Fourier) domain[13,14]
The following examples show both methods:
Filter type Kernel or mask Example

Original Image

Spatial Lowpass

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1684
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470

Spatial Highpass

Pseudo-code:
image = checkerboard
Fourier Representation
F = Fourier Transform of image
Show Image: log(1+Absolute Value(F))

Fourier Low pass

Fourier Highpass

Image padding in Fourier domain filtering

Images are typically padded before being transformed to the Fourier space, the highpass filtered images below
illustrate the consequences of different padding techniques:
Zero padded Repeated edge padded

Notice that the highpass filter shows extra edges when zero padded compared to the repeated edge padding.
Affine transformations
Affine transformations enable basic image transformations including scale, rotate, translate, mirror and shear as
is shown in the following examples:

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1685
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
Transformation Name Affine Matrix Example

Identity

Reflection

Scale

Rotate

where θ = π/6 =30°

Shear

To apply the affine matrix to an image, the image is converted to matrix in which each entry corresponds to the
pixel intensity at that location. Then each pixel's location can be represented as a vector indicating the
coordinates of that pixel in the image, [x, y], where x and y are the row and column of a pixel in the image
matrix. This allows the coordinate to be multiplied by an affine-transformation matrix, which gives the position
that the pixel value will be copied to in the output image.
However, to allow transformations that require translation transformations, 3 dimensional homogeneous
coordinates are needed. The third dimension is usually set to a non-zero constant, usually 1, so that the new
coordinate is [x, y, 1]. This allows the coordinate vector to be multiplied by a 3 by 3 matrix, enabling translation
shifts. So the third dimension, which is the constant 1, allows translation.[15,16]
Because matrix multiplication is associative, multiple affine transformations can be combined into a single affine
transformation by multiplying the matrix of each individual transformation in the order that the transformations
are done. This results in a single matrix that, when applied to a point vector, gives the same result as all the
individual transformations performed on the vector [x, y, 1] in sequence. Thus a sequence of affine
transformation matrices can be reduced to a single affine transformation matrix.
For example, 2 dimensional coordinates only allow rotation about the origin (0, 0). But 3 dimensional
homogeneous coordinates can be used to first translate any point to (0, 0), then perform the rotation, and lastly
translate the origin (0, 0) back to the original point (the opposite of the first translation). These 3 affine
transformations can be combined into a single matrix, thus allowing rotation around any point in the image.

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1686
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
An opening method is just simply erosion first, and then dilation while the closing method is vice versa. In
reality, the D(I,B) and E(I,B) can implemented by Convolution
Structuring
Mask Code Example
element

Use Matlab to read Original image

original = imread('scene.jpg');
image = rgb2gray(original);
[r, c, channel] = size(image);
Original
None se = logical([1 1 1 ; 1 1 1 ; 1 1 1]);
Image
[p, q] = size(se);
halfH = floor(p/2);
halfW = floor(q/2);
time = 3; % denoising 3 times with all method
Original lotus
Use Matlab to dilation
imwrite(image, "scene_dil.jpg")
extractmax = zeros(size(image), class(image));
for i = 1: time
dil_image = imread('scene_dil.jpg');
for col = (halfW + 1): (c - halfW)
for row = (halfH + 1): (r - halfH)
dpointD = row - halfH;
dpointU = row + halfH;
Dilation dpointL = col - halfW;
dpointR = col + halfW;
dneighbor = dil_image(dpointD:dpointU,
dpointL:dpointR);
filter = dneighbor(se); Denoising picture with dilation
extractmax(row, col) = max(filter); method
end
end
imwrite(extractmax, "scene_dil.jpg");
end
Use Matlab to erosion
imwrite(image, 'scene_ero.jpg');
extractmin = zeros(size(image), class(image));
for i = 1: time
ero_image = imread('scene_ero.jpg');
for col = (halfW + 1): (c - halfW)
for row = (halfH +1): (r -halfH)
pointDown = row-halfH;
pointUp = row+halfH;
pointLeft = col-halfW;
Erosion
pointRight = col+halfW;
neighbor =
ero_image(pointDown:pointUp,pointLeft:pointRigh
t);
filter = neighbor(se);
extractmin(row, col) = min(filter);
end
end
imwrite(extractmin, "scene_ero.jpg");
end

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1687
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
Use Matlab to Opening
imwrite(extractmin, "scene_opening.jpg")
extractopen = zeros(size(image), class(image));
for i = 1: time
dil_image = imread('scene_opening.jpg');
for col = (halfW + 1): (c - halfW)
for row = (halfH + 1): (r - halfH)
dpointD = row - halfH;
dpointU = row + halfH;
Opening dpointL = col - halfW;
dpointR = col + halfW;
dneighbor = dil_image(dpointD:dpointU,
dpointL:dpointR);
filter = dneighbor(se);
extractopen(row, col) = max(filter);
end
end
imwrite(extractopen, "scene_opening.jpg");
end
Use Matlab to Closing
imwrite(extractmax, "scene_closing.jpg")
extractclose = zeros(size(image), class(image));
for i = 1: time
ero_image = imread('scene_closing.jpg');
for col = (halfW + 1): (c - halfW)
for row = (halfH + 1): (r - halfH)
dpointD = row - halfH;
dpointU = row + halfH;
Closing dpointL = col - halfW;
dpointR = col + halfW;
dneighbor = ero_image(dpointD:dpointU,
dpointL:dpointR);
filter = dneighbor(se); Denoising picture with closing
extractclose(row, col) = min(filter); method
end
end
imwrite(extractclose, "scene_closing.jpg");
end
Results
Digital image processing deals with manipulation of digital images through a digital computer. It is a subfield of
signals and systems but focus particularly on images. DIP focuses on developing a computer system that is able
to perform processing on an image. The input of that system is a digital image and the system process that image
using efficient algorithms, and gives an image as an output. The most common example is Adobe Photoshop. It
is one of the widely used application for processing digital images.[17]
How it works.

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1688
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
In the above figure, an image has been captured by a camera and has been sent to a digital system to remove all
the other details, and just focus on the water drop by zooming it in such a way that the quality of the image
remains the same.
Libraries involved in Image Processing
The following libraries are involved in performing Image processing in python;
Scikit-image
OpenCV
Mahotas
SimplelTK
SciPy
Pillow
Matplotlib
Scikit-image is an open-source Python package run by the same NumPy members. It uses algorithms and
resources for research, academic and industrial use. It is a simple and straightforward library, even for
newcomers to Python’s ecosystem. The code is high quality, reviewed by peers, and written by a working
community of volunteers.[18]

SciPy is one of Python’s basic science modules (like NumPy) and can be used for basic deception and
processing tasks. In particular, the submodule scipy.ndimage (in SciPy v1.1.0) provides functions that work
on n-dimensional NumPy arrays. The package currently includes direct and offline filtering functions,
binary morphology, B-spline translation, and object ratings.

PIL (Python Imaging Library) is a free Python

programming language library that adds support for
opening, managing, and storing multiple image file
formats. However, its development has stalled, with
the last release in 2009. Fortunately, there is a
Pillow, a PIL-shaped fork, easy to install, works on
all major operating systems, and supports Python 3.
Color-space conversions.

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1689
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
OpenCV (Open Source Computer Vision Library) and applying the same treatment to adequate
is one of the most widely used libraries in computer control images.[20]
programming. OpenCV-Python is an OpenCV
Conclusions
Python API. OpenCV-Python is not only running,
Image processing is a way of doing certain tasks in
because the background has a code written in C / C an image, to get an improved image or to extract
++, but it is also easy to extract and distribute (due
some useful information from it. It is a type of
to Python folding in the front). This makes it a signal processing where the input is an image and
good decision to make computer vision programs
the output can be an image or features/features
more robust.[19,20] associated with that image. Today, image
processing is one of the fastest-growing
technologies. It creates a major research space
within engineering and computer science as well.
Check out the image processing course and
understand the details of the image processing
concept.[20]
Image processing basically involves the following
three steps:
Image import with image detection tools;
Image analysis and management;
The result of which an image or report based on
image analysis can be changed.
There are two types of image processing methods,
namely analogue and digital image processing.
Analogue image processing can be used for hard
copies such as printing and photography. Image
analysts use a variety of translation bases while
using these viewing methods. Digital image
Mahotas is another Python computer and graphics processing techniques facilitate the use of digital
editor. It contains traditional image processing images using computers. The three common
functions such as morphological filtering and categories that all types of data are required to use
functioning, as well as modern computer-assisted when using the digital process are pre-processing,
computational computation functions, including the development, and display of information.[21]
discovery of point of interest and local definitions.
The display is in Python, which is suitable for rapid References
development, but the algorithms are used in C ++ [1] Chakravorty, Pragnan (2018). "What is a
and are fixed at speed. The Mahotas Library is fast Signal? [Lecture Notes]". IEEE Signal
with little code and with little reliance. Processing Magazine. 35 (5): 175–177.
Bibcode: 2018ISPM. . . 35. . 175C. doi:
Image processing is often regarded as improperly 10.1109/MSP.2018.2832195.S2CID 52164353.
exploiting the image in order to achieve a level of
beauty or to support a popular reality. However, [2] Gonzalez, Rafael (2018). Digital image
image processing is most accurately described as a processing. New York, NY: Pearson. ISBN
means of translation between a human viewing 978-0-13-335672-4. OCLC 966609831.
system and digital imaging devices. The human [3] Azriel Rosenfeld, Picture Processing by
viewing system does not see the world in the same Computer, New York: Academic Press, 1969
way as digital cameras, which have additional
sound effects and bandwidth. Significant [4] Gonzalez, Rafael C. (2008). Digital image
differences between human and digital detectors processing. Woods, Richard E. (Richard
will be demonstrated, as well as specific processing Eugene), 1954- (3rd ed.). Upper Saddle River,
steps to achieve translation. Image editing should N. J.: Prentice Hall. pp. 23–28. ISBN
be approached in a scientific way so that others can 9780131687288. OCLC 137312858.
reproduce, and validate human results. This [5] Williams, J. B. (2017). The Electronics
includes recording and reporting processing actions Revolution: Inventing the Future. Springer. pp.
245–8. ISBN 9783319490885.

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1690
 International Journal of Trend in Scientific Research and Development @ www.ijtsrd.com eISSN: 2456-6470
[6] "1960: Metal Oxide Semiconductor (MOS) from the original on 10 June 2016. Retrieved 10
Transistor Demonstrated". The Silicon Engine. October 2019.
Computer History Museum. Archived from the [15] "T. 81 – DIGITAL COMPRESSION AND
original on 3 October 2019. Retrieved 31
CODING OF CONTINUOUS-TONE STILL
August 2019. IMAGES – REQUIREMENTS AND
[7] James R. Janesick (2001). Scientific charge- GUIDELINES" (PDF). CCITT. September
coupled devices. SPIE Press. pp. 3–4. ISBN 1992. Archived (PDF) from the original on 17
978-0-8194-3698-6. July 2019. Retrieved 12 July 2019.
[8] Boyle, William S; Smith, George E. (1970). [16] "The JPEG image format explained". BT. com.
"Charge Coupled Semiconductor Devices". BT Group. 31 May 2018. Archived from the
Bell Syst. Tech. J. 49 (4): 587–593. original on 5 August 2019. Retrieved 5 August
doi:10.1002/j.1538-7305.1970.tb01790.x. 2019.
[9] Fossum, Eric R. (12 July 1993). "Active pixel [17] "What Is a JPEG? The Invisible Object You
sensors: Are CCDS dinosaurs?". In Blouke, See Every Day". The Atlantic. 24 September
Morley M. (ed.). Charge-Coupled Devices and 2013. Archived from the original on 9 October
Solid State Optical Sensors III. Proceedings of 2019. Retrieved 13 September 2019.
the SPIE. Vol. 1900. pp. 2–14. Bibcode:
[18] Baraniuk, Chris (15 October 2015). "Copy
1993SPIE. 1900. . . . 2F. CiteSeerX 10. 1. 1. protections could come to JPEGs". BBC News.
408. 6558.
Archived from the original on 9 October 2019.
doi:10.1117/12.148585.S2CID10556755. Retrieved 13 September 2019.
[10] Fossum, Eric R. (2007). "Active Pixel Sensors".
[19] Grant, Duncan Andrew; Gowar, John (1989).
S2CID 18831792.
Power MOSFETS: theory and applications.
[11] Matsumoto, Kazuya; et al. (1985). "A new Wiley. p. 1. ISBN 9780471828679. The metal-
MOS phototransistor operating in a non- oxide-semiconductor field-effect transistor
destructive readout mode". Japanese Journal of (MOSFET) is the most commonly used active
Applied Physics. 24 (5A): L323. Bibcode: device in the very large-scale integration of
1985JaJAP. . 24L. 323M. digital integrated circuits (VLSI). During the
doi:10.1143/JJAP.24.L323.S2CID108450116. 1970s these components revolutionized
electronic signal processing, control systems
[12] Fossum, Eric R.; Hondongwa, D. B. (2014). "A
and computers.
Review of the Pinned Photodiode for CCD and
CMOS Image Sensors". IEEE Journal of the [20] Shirriff, Ken (30 August 2016). "The
Electron Devices Society. 2 (3): 33–43. Surprising Story of the First Microprocessors".
doi:10.1109/JEDS.2014.2306412. IEEE Spectrum. Institute of Electrical and
[13] "CMOS Image Sensor Sales Stay on Record- Electronics Engineers. 53 (9): 48–54.
doi:10.1109/MSPEC.2016.7551353. S2CID
Breaking Pace". IC Insights. 8 May 2018.
Archived from the original on 21 June 2019. 32003640. Archived from the original on 13
October 2019. Retrieved 13 October 2019.
Retrieved 6 October 2019.
[21] "1979: Single Chip Digital Signal Processor
[14] Ahmed, Nasir (January 1991). "How I Came
Up With the Discrete Cosine Transform". Introduced". The Silicon Engine. Computer
History Museum. Archived from the original on
Digital Signal Processing. 1 (1): 4–5.
3 October 2019. Retrieved 14 October 2019.
doi:10.1016/1051-2004(91)90086-Z. Archived

@ IJTSRD | Unique Paper ID – IJTSRD51728 | Volume – 6 | Issue – 5 | July-August 2022 Page 1691