

FINAL YEAR PROJECT

PROPOSAL ON

IMAGE ENCRYPTION AND DECRYPTION USING ADVANCED CRYPTOGRAPHIC

ALGORITHM

PRESENTED BY

OMONIYI WARIZ OLOLADE

CSC/2020/1149

SUBMITTED TO SUPERVISOR:

DR. AWOYEMI

IN PARTIAL FUFILLMENT OF THE REQUIREMENTS FOR THE AWARD OF

BACHELOR OF COMPUTER SCIENCE (B.Sc COMPUTER SCIENCE)

DATE: NOVEMBER, 2024

 

INTRODUCTION
The security and privacy of sensitive information, including images, have become incredibly important as technology becomes a bigger part of
our daily lives. With individuals, organizations, and governments sharing massive amounts of visual data online, concerns about unauthorized
access, tampering, and interception have grown significantly. Because of this, there’s now a stronger focus on using solid security measures to
protect the confidentiality and integrity of this data (Chanal et al., 2020). Image encryption and decryption—key components of modern
cryptography—play a critical role in addressing these issues.

Cryptography, the science behind secure communication, involves techniques for turning information into unreadable formats. Only those with
the right decryption keys can convert this data back to a readable form. When it comes to images, encryption works specifically on visual data
like photos, diagrams, and medical scans. It transforms them into encrypted formats that are nearly impossible to interpret without the correct
keys.

This process ensures that even if someone gains unauthorized access to the encrypted images, they can’t understand the content without the
decryption key. Encryption doesn’t just protect stored images—it also safeguards images shared over digital networks, making it much harder for
attackers to intercept or misuse the data being transmitted (Alkhalil et al., 2021).

Take medical imaging, for example. Hospitals often share sensitive images, such as X-rays or MRIs, within their departments or with specialists
for diagnosis. Since these images carry private patient information, maintaining their confidentiality is crucial. Image encryption ensures that only
authorized medical professionals, equipped with the right decryption keys, can access and interpret these images, keeping patient data safe from
unauthorized access.

Similarly, industries like e-commerce and finance rely on image encryption to protect financial transactions and customer data. Social media
platforms and other websites handling user-generated content also use encryption to shield users’ images from malicious actors. One widely used
technique in image security involves applying cryptographic algorithms like AES (Advanced Encryption Standard) to pixel data. These
algorithms are designed with complex mathematics to ensure they’re computationally secure, making it nearly impossible for unauthorized
parties to reverse-engineer the encryption and retrieve the original image.

By securing image data, these encryption methods lay the groundwork for scalable solutions that can be applied in real-world contexts, like
secure image-sharing platforms or systems for confidential communication.

AIM AND OBJECTIVES

This project is focused on analyzing and improving image encryption and decryption techniques. It aims to identify vulnerabilities in current
algorithms, develop more efficient solutions, and explore innovative approaches to ensure secure transmission and storage of visual data in
today’s rapidly evolving digital landscape. The system developed will adhere to the following essential criteria:

I. Robust Security: The encrypted images should be highly resistant to unauthorized access and various cryptanalysis techniques.
II. Efficiency: The system must enable rapid encryption and decryption, even for large, high-resolution images, without compromising
performance.
III. Scalability: It should be capable of handling diverse image types, ranging from small grayscale images to complex, high-resolution
color images.
IV. Practical Usability: The solution should be user-friendly and seamlessly integrable with existing digital systems, such as secure file-
sharing platforms and online storage services.

OBJECTIVES
The study aims to design an advanced cryptographic framework tailored to image data by integrating symmetric encryption for speed and
asymmetric encryption for enhanced security. It also involves implementing the cryptographic system using optimized algorithms such as AES
and leveraging tools like Python to create and refine the system. Another key objective is evaluating the system’s performance through
established metrics to assess improvements in speed, security, and usability compared to existing solutions. Finally, the project emphasizes
optimizing scalability to ensure the system efficiently processes various image formats, including JPEG, PNG, and BMP, without compromising
performance or security.

STATEMENT OF PROBLEM
In the modern digital landscape, safeguarding image data has become crucial due to the frequent exchange of images across various platforms.
While traditional encryption techniques work well for text, they often fall short when dealing with the complex nature and larger size of image
files. The growing threats of cyberattacks, particularly in fields like healthcare and multimedia, highlight the need for more sophisticated methods
to secure visual data.
 

This research explores innovative strategies for image encryption and decryption, aiming to address the limitations of current techniques. By
tackling the unique challenges of securing pixel-based image data, the study seeks to contribute to the development of more secure and efficient
encryption methods.

Key Issues:
i. Computational Overhead: Standard cryptographic algorithms, when applied to large images, result in significant processing
delays and resource consumption, making them impractical for real-time applications.
ii. Vulnerability to Attacks: Simple encryption schemes are often susceptible to brute-force attacks, differential cryptanalysis,
and statistical analysis due to the patterns present in image data.
iii. Usability Challenges: Existing systems often lack user-friendly interfaces and integration capabilities, limiting their
adoption in everyday use cases, such as secure image sharing.
iv. Scalability Concerns: Many encryption algorithms struggle with scalability when handling diverse image types,
resolutions, and formats.

OVERVIEW OF CURRENT SYSTEMS

Current methods for securing images rely on two main approaches: traditional cryptographic algorithms and specialized techniques tailored
specifically for image encryption. Each approach has its own strengths and limitations.

Traditional Cryptographic Algorithms are commonly used for securing data, including images, but they often face challenges when applied to
large or complex image files:

i. AES (Advanced Encryption Standard) is widely recognized for its strong security. However, encrypting large image data with AES
can be computationally intensive, resulting in processing delays that may hinder performance in real-time applications.

ii. RSA (Rivest-Shamir-Adleman) provides robust asymmetric encryption, making it ideal for secure key exchanges. Yet, it is inefficient
for encrypting large image files because of its reliance on computationally heavy operations.

iii. ECC (Elliptic Curve Cryptography) offers a lightweight alternative to RSA, delivering strong security with lower resource
requirements. Despite its efficiency, ECC is less commonly used for image encryption, potentially limiting its practical applications in
this domain.

iv. Specialized Image Encryption Techniques, on the other hand, are tailored to the unique properties of visual data and offer alternative
methods for securing images:

v. Pixel Shuffling rearranges the positions of pixels in an image to obscure its content. While this method is fast and easy to implement,
it does not provide sufficient security when used alone and must be combined with other encryption techniques for effectiveness.

vi. Chaotic Systems utilize principles from chaos theory to generate random encryption keys. These systems are highly secure but present
challenges in terms of implementation and optimization, making them less accessible for widespread use.

vii. Both approaches highlight the trade-off between security, efficiency, and usability, underscoring the need for hybrid solutions that
leverage the strengths of each method to address their limitations.

OVERVIEW OF THE PROPOSED SYSTEM

The proposed system seeks to address the limitations of current methods by combining cryptographic algorithms with image-specific
optimizations and steganographic techniques. This approach is designed to deliver high levels of security, efficiency, and adaptability, making it
suitable for various applications.

The system employs a hybrid encryption approach that merges symmetric encryption for faster processing with asymmetric encryption to
securely manage key exchanges. Dynamic key generation is also introduced to bolster resistance against cryptographic attacks. Additionally, the
system incorporates image-optimized encryption, leveraging the unique structure of images, such as pixel correlation, to reduce computational
demands while maintaining image quality during decryption, ensuring minimal loss of fidelity.
 

Steganographic integration is another key aspect, enabling the concealed transmission of sensitive images over unsecured channels. The system is
also highly scalable and flexible, supporting various image formats like JPEG and PNG, as well as different resolutions. Furthermore, it is
designed for compatibility with cloud platforms and mobile devices, allowing real-time image sharing and secure storage.

The workflow of the proposed system begins with the encryption process, where the input image is pre-processed—compressed if necessary—to
reduce size and redundancy. Optionally, the encrypted image can be embedded within a cover image using steganography for added security. In
the decryption process, the encrypted image is received by the intended recipient, who securely retrieves the decryption keys through asymmetric
encryption protocols. Finally, the original image is restored with minimal degradation, ensuring that the quality and fidelity of the visual data
remain intact.

IMPLEMENTATION STEPS
The implementation of the proposed system will follow a carefully structured approach to achieve the desired objectives.

The process begins with Requirement Analysis, where the specific cryptographic needs for securing images are identified. This includes defining
the target image formats, sizes, and resolutions, as well as establishing performance requirements like encryption speed and system scalability.

Next is the Algorithm Design phase. A hybrid cryptographic framework will be developed, combining AES (Advanced Encryption Standard) for
encryption with ECC (Elliptic Curve Cryptography) for secure key exchange. Additionally, a steganographic module will be created to embed
encrypted images into cover images. Adaptive algorithms will also be implemented to optimize processing for different image types and
resolutions.

During the System Development stage, Python will be used for programming, and a user-friendly graphical interface will be built to make the
system accessible and easy to use.

The system will then undergo Testing and Evaluation, focusing on functionality, performance, and security. Functionality tests will ensure images
are encrypted and decrypted accurately, while performance tests will measure encryption and decryption speeds, particularly for high-resolution
images. Security tests will evaluate the system’s resilience to attacks such as brute force, differential cryptanalysis, and statistical attacks.

After testing, the Optimization phase will involve fine-tuning the algorithms to strike the right balance between speed and security. Parallel
processing techniques will be implemented to enhance performance when handling large datasets, and the system will be adjusted for
compatibility with low-resource environments.

Finally, the system will move to Deployment and Integration, where it will be tested in real-world applications such as secure image-sharing
platforms. It will also be integrated with cloud services to enable secure storage and retrieval of encrypted images, ensuring it meets practical
usability requirements.

CONTRIBUTIONS TO KNOWLEDGE
This research makes significant contributions to the field of image encryption by introducing a novel key generation technique based on Gaussian
distribution. This innovative approach provides a fresh perspective on encryption, potentially improving security measures for protecting visual
data. Furthermore, the development of a practical desktop application using Python and Tkinter offers a functional tool for managing images
securely.

A noteworthy aspect of this study is the use of Mean Squared Error (MSE) as a performance metric, which provides a quantitative measure of the
system’s effectiveness in preserving image quality. Together, these contributions push the boundaries of current understanding and applications
in image encryption technology.

The research also introduces a novel hybrid cryptographic framework that blends the speed of symmetric encryption algorithms like AES with the
robust security of asymmetric algorithms such as RSA or ECC. This framework is specifically optimized for image data, delivering a tailored
solution for protecting visual content.

Incorporating steganography enhances the system’s capabilities by allowing sensitive images to be transmitted discreetly, even over unsecured
communication channels. The system also emphasizes performance, demonstrating faster encryption and decryption times that make it viable for
real-time applications. High-quality image restoration post-decryption ensures minimal fidelity loss, reflected in high PSNR (Peak Signal-to-
Noise Ratio) values.

Designed as a scalable and practical solution, the system can handle a wide variety of image formats, resolutions, and sizes. Its compatibility with
mobile devices further enhances its usability, making it adaptable to diverse user needs and platforms.

Beyond its immediate applications, this research contributes to the broader field of cryptography and image security. The methodologies and
findings will provide a foundation for future studies, potentially inspiring encryption solutions for other types of multimedia data, including
videos and 3D models.
 

Real-world applications of this system are wide-ranging. It can be integrated into secure communication platforms, online storage systems, and
tools for sharing confidential documents. Industries such as healthcare, defense, and social media stand to benefit from the robust image security
solutions this project delivers, addressing critical challenges in safeguarding sensitive visual information.

LITERATURE REVIEW

S/ Title Authors & Abstract Methodology Weakness/

N Year Limitation

i. Image Kolivand, This paper The study The study

encryption presents an uses various identifies
H,
techniques: A exhaustive methodologi several
comprehensive Hamood,
review of es for limitations. One
review.
SF, research encrypting major issue is
Multimedia
Tools and Asadianfam within the images, the trade-off
Applications field of focusing on between
, S and
image chaos-based computational
Rahim, MS encryption approaches. efficiency and
(2024) techniques. It The research encryption
commences categorises strength, with
with a encryption robust methods
general techniques often being
introduction into spatial, resource-
to image frequency, intensive.
encryption, and hybrid Certain
providing an domains, techniques, such
overview of providing a as quadtree
the comprehensi scrambling, are


fundamentals ve survey of vulnerable to

. existing attacks like the
Subsequently methods. "jigsaw
, it explores a Chaos-based conundrum"
comprehensi encryption is and "neighbour
ve highlighted ambush." Some
exploration for its use of frameworks,
of chaos- randomised particularly
based image key systems those relying on
encryption, and its phase spectra,
encompassin application are
g various in both full incompatible
methods and and selective with image
approaches encryption compression
within this frameworks. standards,
domain. Full limiting their
These encryption practical
methods processes all applications.
include full pixels Additionally,
encryption uniformly, despite
techniques as while advancements,
well as selective chaos-based
selective encryption encryption
encryption targets systems
strategies, specific sometimes lack
offering regions for sufficient
insights into computation robustness
their al efficiency. against
principles Techniques sophisticated
and such as tent cryptanalysis.
 

applications. maps and Inconsistencies

The authors SCAN-based in their
place sig methods in implementation
nificant the spatial can create
emphasis on domain and security
surveying phase spectra vulnerabilities.
prior techniques in
research the
contributions frequency
, shedding domain are
light on among the
noteworthy key
development approaches
s within the discussed.
field. Hybrid
Additionally, methods
the paper combining
addresses spatial and
emerging frequency
challenges elements aim
and issues to enhance
that have security
arisen as a further
consequence
of these
advancement
s.
ii A Alyaa I. Today, with It classifies Techniques
Comprehensive Dawood 1, the encryption relying on low-
continuous approaches dimensional
Review of Qabeela Q.
 

Color Image Thabit, increase in based on chaotic maps

Encryption Taqwa O. the use of their may have
computer algorithms, limited key
Technology Fahad
networks and such as spaces, making
2023. the rapid chaotic them vulnerable
evolution of maps, DNA to brute-force
information computing, attacks.
technologies. and other Methods
The secure advanced incorporating
transmission systems. The DNA-based or
of data over paper deep learning
the Internet evaluates approaches,
has become these while robust,
an urgent methods in often suffer
necessity to terms of their from high
preserve the security computational
privacy of features, costs, limiting
users and efficiency, their practicality
protect and for real-time
sensitive resistance to use. Chaotic
information attacks. map-based
from theft Some algorithms can
and approaches, be highly
distortion. like chaotic sensitive to
images are systems, initial
most of this employ 2D conditions,
transferred and 3D where minor
data, so it chaotic maps errors may
was for pixel compromise
necessary to scrambling security.


protect it by and data Furthermore,

encrypting diffusion, some methods
them using reducing lack scalability
algorithms correlation or compatibility
that ensure between with large
the image pixels. datasets, leading
protection of DNA-based to increased
information methods encryption and
access to the leverage decryption
receiver. biological times.
Color images principles for
contain encoding and
sensitive scrambling
information matrices,
and details while deep
that must be learning
secured and models are
protected. used to
This paper predict
produces a chaotic
comprehensi sequences,
ve review of enhancing
image encryption
encryption strength.
methods and Additionally,
classifies the paper
them based compares
on various these
concepts techniques
such as using
 

chaotic parameters
maps, DNA, such as key
etc. with space size,
comparisons computation
between al efficiency,
existing and
approaches vulnerability
to accessing to different
different types of
security attacks.
parameters.
Additionally,
the types of
encryption
keys were
reviewed
along with
some
common
types of
attacks and
the most
important
methods for
measuring
encryption
efficiency.
iii Security of Saba Inam , The Internet The research The
End-to-End Shamsa of Medical introduces a performance of
medical images Kanwal, Things deep the encryption
 

encryption Anousha (IoMT) learning- and decryption

system using Anwar, connects based network is
trained deep Noor medical encryption influenced by
learning Fatima devices and and the quality and
encryption and Mirza, wearables to decryption diversity of the
decryption Hessa enhance network training data;
network Alfraihi healthcare, designed to insufficient or
necessitating safeguard non-
encryption to medical representative
protect images data may lead to
sensitive throughout suboptimal
patient data. their results.
This study transmission. Additionally,
proposes The the
deep methodology computational
learning- involves complexity of
based training a the deep
encryption deep neural learning model
for medical network to can be a
images using perform both concern,
Cycle_GAN, encryption potentially
a Generative and requiring
Adversarial decryption significant


Network. tasks, resources for

Cycle GAN ensuring that training and
enables the deployment.
image encrypted The study also
transformatio images notes that while
n without remain the proposed
requiring confidential system
paired and can only demonstrates
training data, be accurately high security,
addressing decrypted by its resilience
challenges in authorised against
obtaining entities. The advanced
annotated study cryptographic
medical employs a attacks remains
datasets. It dataset of an area for
uses a medical further
generator to images, investigation.
modify applying the Moreover, the
images and a trained generalisability
discriminator network to of the model to
to assess its different types
distinguish effectiveness of medical
real from in images and its


fake, with a maintaining scalability in

Binary-Cross image real-world
Entropy loss quality and applications are
function for security. aspects that
training. warrant
Experiments additional
on skin research.
cancer
datasets
show this
method
delivers
efficient,
systematic,
and secure
encryption
compared to
other modern
approaches,
ensuring
confidentialit
y and
authenticity.
 

iv Image Mr. Image The The main

Cryptography Shivdatta protection is methodology limitation or
(Encryption Pawar, vital as it for Image weakness
and Mrs. dominates Cryptograph identified in the
Decryption) Surekha multimedia y approach is the
Kohle data. With (Encryption efficiency of the
2022. increasing and algorithm.
digital image Decryption) Specifically, the
transmission as presented decryption
and storage, in the journal process is
ensuring by Mr. significantly
confidentialit Shivdatta slower than
y, integrity, Pawar and encryption.
and Mrs. Surekha While
authenticity Kohle encryption takes
has become focuses on less time,
crucial. transforming decryption is
Image image data roughly four
encryption into a matrix times slower,
addresses format. The which can be an
this by process issue in time-
scrambling begins by sensitive
pixels and converting applications.


reducing the image Additionally,

pixel into a the complexity
correlation to grayscale of the process,
enhance representatio involving
security. n by matrix
This paper averaging its manipulations
reviews RGB values. like Gaussian
various Gaussian elimination,
image elimination introduces
encryption with partial computational
techniques, pivoting and overhead,
discussing row making it less
their exchanges is efficient
methods, applied to compared to
advantages, this image simpler
and matrix, encryption
limitations, generating algorithms.
while an upper
providing an triangular
overview of matrix that
image serves as the
encryption's encrypted
significance. form. The
 

decryption
process uses
the
decryption
key
generated
during
encryption to
reverse the
transformatio
n and
retrieve the
original
image.
v Image Yousef, Ensuring the A Survey of as discussed in
Encryption Alghamdi, confidentialit Design and the paper,
Algorithms: A and y and Evaluation includes the
Survey of ArslanMuni privacy of Metrics" by inherent
Design and r images Yousef vulnerabilities
Evaluation 2024. shared over Alghamdi associated with
Metrics unsecured and Arslan some chaotic-
networks is Munir based
vital. provides an encryption


Common extensive methods, such

encryption overview of as the limited
methods like various range of chaotic
AES, image behavior and
Twofish, and encryption susceptibility to
RSA, while algorithms chosen-plaintext
secure, are and their attacks. For
unsuitable correspondin instance, while
for real-time g evaluation chaotic systems
image metrics. The like the Logistic
encryption methodology and Arnold
due to their adopted in maps are fast
slow speeds the paper and effective,
and high includes an they can be
computation in-depth prone to
al demands. classification reverse-
As a result, of image engineering
researchers encryption under specific
have techniques conditions,
developed based on the which may
specialized strategies compromise
algorithms they employ, security.
for image such as


encryption. chaotic
This paper maps,
reviews and permutation-
classifies substitution
these schemes, and
algorithms key
into seven management
approaches methods.
based on The authors
their evaluate
techniques. It these
evaluates algorithms
each based based on
on security, their
quality, and security,
efficiency computation
metrics, al efficiency,
providing and quality
upper and of encrypted
lower images,
bounds for making
these comparisons
evaluations. between
 

The different
discussion techniques to
also covers guide future
the strengths research and
and application
weaknesses in this field
of various
methods and
their
suitability
for different
applications
vi Fast and Secure Rasha S. The rapid In this the paper lacks
Image Ali, Maha evolution of article, we a
Encryption Khalil network aim to utilize comprehensive
System Ibrahim, technologies, the security
Using New Saad N. where lightweight analysis. It does
Lightweight Alsaad computer Trivium not address
Encryption 2024. vision and algorithm to potential
Algorithm the Internet be suitable vulnerabilities
seamlessly for securing to cryptographic
merge, poses transmitted attacks like
significant data on the differential


challenges in networks. analysis,

safeguarding Our main chosen-plaintext
end-user contributions attacks, or
data. To are as known-plaintext
address the follows: attacks, the
growing 1. Proposed paper does not
need for a compact provide detailed
secure data synchronous information on
in smart stream key
environment cipher management
s, system with practices, such
lightweight improved as key
and efficient parameters generation,
security that can distribution,
solutions are offer higher storage, and
in high security and sensitivity
demand. better chip analysis, the
Notably, the size and paper does not
Trivium, a power provide detailed
well-known consumption benchmarks or
encrypt performance. comparisons
stream 2. Outline with other
cipher the lightweight


(eSTREAM) guidelines encryption

project, has for selecting algorithms.
been the effective
target parameters
of several when using
attacks. This chaotic
paper functions.
proposes an
innovative
approach for
fast and
highly
secured
image
encryption
using byte
scrambling
and a
modified
version of
the
Trivium
algorithm.


The Henon
map is
utilized to
generate
random
numbers for
permutation,
reducing
time and
increasing
security.
Performance
is measured
on 100
colour
images with
different
several tests.
The
results
present a fast
and robust
security


system for
the
transmission
process. It is
more
efficient than
the
traditional
Trivium
method. The
encrypted-
decryption
time is
reduced by
1:24 times,
making it a
quick
security
system for
surgical
telepresence
and
multimedia
 

visuals.

vii A N. Mahendi The The paper Limited focus

reviews on real-time
Comprehensive ran, explosive
methods for implementation
Analysis C. Deepa generation of encrypting s of hybrid
and methods in
on Image 2021. data in the
compressing practical
Encryption digital era images, systems,
focusing on computational
and Compressi directs to the
hybrid costs and
on Techniques requirements techniques challenges with
that balance real-time use,
with the Assess of effective
security and especially in
ment approaches efficiency. It low-resource
evaluates systems.
of Performance to transmit
these
Evaluation and store the approaches
using metrics
Metrics data. The
like
necessities of compression
ratio and
security and
encryption
limited strength.
resources
have led to
the
development
of image
encryption
 

and
compression
approaches,
respectively.
The digital
contents are
transmitted
over the
internet
which may
subject to
security
threats. To
overcome
these
limitations,
image
encryption
approaches
are
developed.
vii Multimedia KHALID This survey The study While the paper
M. reviews how provides a
i Security Using aims to
HOSNY, encryption comprehensive
 

Encryption: A (Senior review the secures review, it

Member, multimedia highlights a few
Survey state of
IEEE), like images, issues. Many
MOHAME secure and videos, and studies use
D A. audio. It inconsistent
privacy-
ZAKI,, categorizes metrics, making
NABIL A. preserving different it hard to
LASHIN, encryption compare
encryption
MOSTAFA methods methods fairly.
schemes based on Some focus
M. FOUD,
their use and only on basic
applicable to
(Senior compares encryption
digital their techniques,
Member,
performance ignoring recent
multimedia,
IEEE), against advancements.
such as common The lack of
AND
threats like real-world
digital
HANAA hacking or applications and
images, tampering. practical
M. HAMZ
The authors examples is
digital video,
2023. assess these another
and digital methods limitation.
using metrics Additionally,
audio. An
like security the survey
extensive strength and doesn’t fully
the quality of explore how to
analysis of
encrypted integrate
the existing content. encryption with
They also other security
cryptography
highlight methods like
schemes and potential watermarking
attacks, such or
multimedia
as brute- steganography.
encryption force or Finally, though
statistical future directions


algorithms analysis, and are mentioned,

explore how newer
will be
well these technologies
conducted to encryption like quantum-
techniques safe encryption
give an
withstand aren’t deeply
extensive them. Lastly, discussed.
the paper
overview of
outlines
the current areas for
improvement
state of
to guide
security future
research and
encryption
development
schemes of more
robust
specifically
multimedia
designed for security
systems.
digital
multimedia
technology.
The survey
results will
be used to
understand
better the
effectiveness
and
 

reliability of
secure
multimedia
encryption
schemes and
contribute to
developing
efficient and
secure
encryption
schemes in
the future.
ix SPN based Muhammad This research This study While the
Sajjada, paper proposes a method is
RGB image
Tariq Shah, proposes a new way to innovative, it
encryption over Tanveer ul novel encrypt RGB has some
Haqa, approach for images using drawbacks. The
Gaussian
Bander constructing a structure use of complex
integers Almutairi, substitution called mathematical
Qin Xin boxes (S- Substitution- operations can
2024. boxes) over Permutation make it slow on
Gaussian Network devices with
integers, (SPN). The limited
which are process uses processing
complex Gaussian power. The
numbers integers (a study focuses
with integer type of more on theory
coefficients. mathematical than on testing
The system) to the method in


proposed create more real-world

method is complex and situations or on
based on the secure larger images.
properties of encryption. Additionally,
the Gaussian The method the method
integers and transforms hasn’t been
their image data integrated with
arithmetic by other useful
operations substituting techniques, like
and ensures and shuffling compressing or
the S- pixel values, watermarking
boxes exhibit making it images, which
strong harder to are often needed
cryptographi predict or in practical
c properties. reverse the applications.
Furthermore, encryption. It Scalability and
the paper also includes adaptability to
demonstrates a secure key different
how these S- generation multimedia
boxes can be system to formats also
utilized for protect remain
image against unaddressed
encryption brute-force
through a attacks. The
substitution- researchers
permutation tested the
network method’s
(SPN) security and
over effectiveness
by
Gaussian
evaluating
integers. how well it
resists
various
attacks and
 

how
unpredictabl
e the
encrypted
data is.

x IMAGE Shaolin Encryption is This paper While some

Kataria, a key tool for reviews methods are
ENCRYPTION
Elio Jordan securing different secure, older
TECHNIQUES Lopes, multimedia methods for ones like DES
Bevis data as it encrypting are outdated and
AND
Mathew protects files images, vulnerable to
COMPARATI Niravil, like images, focusing on modern threats.
Shashank text, and popular Techniques like
VE
Keshav audio while algorithms RSA, while
ANALYSIS 2021. they are like RSA, effective, are
transmitted AES, and computationally
online. By DES, as well heavy and
scrambling as more unsuitable for
the data and advanced real-time use.
using an techniques The study
encryption such as mostly
key, DNA-based evaluates
unauthorized encryption security but
access is and chaotic doesn’t dive
prevented, systems. It into other
ensuring that compares important
only the these factors like
intended approaches energy use or
recipient based on storage
with the their speed, demands.
correct security, and Additionally,
decryption ability to advanced
key can resist attacks. techniques often


access the The authors need high-end

original also suggest hardware,
information. an improved limiting their
The version of practicality in
encryption the AES everyday
process algorithm to settings.
replaces or address
transforms common
data into an issues, like
unrecognizab ensuring
le format, encrypted
making it images don’t
secure reveal visual
during clues about
transmission. the original
The goal of data
this project is
to study and
understand
key
encryption
and
decryption
methods,
focusing on
popular
algorithms
like DES,
AES, and
RSA. This
includes
examining
how these
algorithms
work, their
 

application
to different
types of data,
and
evaluating
their
performance
and
efficiency
under
various
conditions.

xi An Image Cong Xu, In this paper, The image While the

encryption algorithm offers
Encryption Jingru Sun we propose
algorithm strong
Algorithm and an image developed by encryption, it
Cong Xu and has some
Based on Chunhua encryption
colleagues weaknesses. It
Random Walk Wang algorithm uses a might be
combination vulnerable to
and Hyper- 2019. based on
of random more advanced
chaotic random walk walk attacks, such as
techniques brute-force or
Systems. and two
and differential
hyper- hyperchaotic cryptanalysis,
systems to because of its
chaotic
improve reliance on
systems. The security. The chaotic systems.
algorithm Additionally,
random walk
first the randomness
method is scrambles introduced by
the positions the random
adopted to
of image walk method


scramble the pixels using could become

random walk predictable in
position of
methods, some scenarios,
pixels within making it which could
harder to lower its
a block.
predict the effectiveness
Furthermore, image for certain types
structure. of images. The
the
Then, it uses method also
permutation chaotic demands a lot
systems of
operation
(Lorenz and computational
between Chen power due to
systems) to the complexity
blocks is
generate key of the chaotic
presented to sequences systems and
that further multiple
enhance the
enhance the encryption steps
scramble encryption. involved,
The method making it less
effect. Thus,
also relies on practical for
high generating devices with
initial keys limited
correlation
through an resources
among pixels MD5 hash of
the image,
of original
which are
image is applied to
control the
broken by
chaos in the
permutation. encryption
process. The
Moreover,
system is
the chosen tested
against
 

plaintext common
attacks to
attack is used
evaluate its
to test the robustness
anti-attack
ability of the
proposed
algorithm.
By analyzing
experimental
results and
comparing
with other
image
encryption
algo- rithms,
we show that
the proposed
algorithm
has better
performance
and higher
security.
xii A New Image WANG JI This paper The One potential
introduces an approach limitation of
 

Encryption JUN AND efficient and centers on this encryption

secure image creating a scheme could
Algorithm TAN SOO
encryption single S-box be its reliance
Based on FUN, scheme using a low- on the logistic
based on a dimensional map, which,
Single S-Box (Member
chaotic chaotic while efficient,
and Dynamic IEEE) system and system, might not
S-boxes. specifically a provide as high
Encryption 2021.
Unlike logistic map. a level of
Step previous This reduces unpredictability
methods that the as more
used computation complex chaotic
multiple or al systems.
double S- complexity Additionally,
boxes (which compared to the single S-box
were high- construction,
vulnerable to dimensional though efficient,
attacks), the chaotic may still face
authors systems limitations in
propose commonly handling very
using a used in large or high-
single 10 × previous entropy images,
26 S-box methods. potentially
created from The reducing its
a low- algorithm effectiveness in
dimensional incorporates certain high-
chaotic a dynamic security
system. This encryption contexts
approach step to
avoids address the
complex correlation
math while between
providing image pixels,
improved enhancing
efficiency security.


and higher Pixel

entropy, diffusion is
which means employed to
more secure further
encryption. disrupt the
Additionally, statistical
the paper properties of
introduces a the image,
dynamic making it
encryption more
step method resistant to
to address differential
issues like attacks. The
pixel S-box size is
correlation extended to
and 10×26 to
determinism handle more
that could complex
weaken encryption
multiple S- applications
box methods. without
This new compromisin
method g security or
ensures increasing
better pixel algorithmic
randomness, complexity
enhancing
both the
security and
efficiency of
image
encryption in
real-world
applications.
 

xii A new four‐tier Khalid M. This paper The The limitations

proposes a methodology of this approach
i technique for Hosny and
four-tier involves a include its
efficient Sara T. encryption four-step computational
technique for approach: complexity,
multiple Kamal
securely combining especially with
images 2024. transmitting images, large datasets,
multiple randomizing and potential
encryption
digital their order, vulnerabilities
images. The scrambling in certain
method aims them through advanced
to enhance various attacks.
both security patterns, and However, it
and diffusing the shows good
transmission scrambled effectiveness in
speed. The image using security against
process an altered common threats
involves: chaotic map.
Creating an
Augmented
Image,
Randomizing
the Image,
Scrambling
the Image,
Diffusing the
Image.
The paper
provides a
flowchart,
pseudo-code,
and an
example to
 

make the
method easy
to
understand.
Through
experiments,
it shows that
this four-tier
encryption
technique is
highly
effective and
secure
against
various
attacks,
making it a
promising
solution for
image
security in
transmission.

xi Computational Mandeep Images often Categorized No unified

Image Kaur, contain techniques framework for
v
Encryption Surender sensitive by comparing
Techniques: A Singh, information, foundational techniques,
Comprehensive Manjit making their concepts, Limited
Review Kaur 2021. protection Evaluated adaptability to
crucial, performance diverse
especially in using metrics applications,
areas like like Scalability
military encryption issues with
communicati speed and large datasets,
on, remote resistance to Vulnerability to


sensing, and attacks, advanced future

medical Provided threats, High
imaging. examples resource
Encryption is and demands for
a widely- comparisons some methods.
used method to highlight
for securing their
such data. applications.
This paper
reviews
various
image
encryption
techniques,
categorizing
them based
on
approaches
like chaotic
maps, DNA
coding,
compressive
sensing, and
optical
methods. It
compares
these
methods
using
security
performance
metrics and
highlights
potential
areas for
 

improvement
to inspire
further
research in
the field of
image
encryption.

xv Design of Daihan Xu, This paper The The integration

proposes an methodology of hyperchaos
artificial Guodong
image involves and compressed
intelligence Li, Wenxia encryption converting sensing
algorithm an image increases
image Xu,
designed for into a sparse computational
encryption Chengjing artificial form using complexity,
intelligence Discrete making the
algorithm Wei 2023.
(AI) Wavelet algorithm less
based on environment Transform suitable for real-
s, focusing (DWT), time
hyperchaos
on privacy which applications or
protection. facilitates devices with
The method efficient limited
starts by processing. hardware
converting An improved capabilities. The
the image hyperchaotic reliance on
into a sparse system is advanced
format using then hardware may
Discrete employed to limit its
Wavelet create highly accessibility,
Transform randomized and scalability
(DWT). It encryption remains a
then uses an keys and concern when
enhanced introduce a handling large
hyperchaotic dual reset datasets or
 

sequence, mechanism. continuous

which Circular image streams.
incorporates diffusion Further
a dual reset operations optimization is
mechanism further necessary to
and a scramble the address these
circular image’s challenges and
diffusion pixel enhance its
process to arrangement. practicality.
further Finally,
scramble the compressed
image data. sensing is
Lastly, applied using
compressed a random
sensing is Gaussian
applied, measurement
where a matrix,
random which
Gaussian simultaneous
matrix is ly reduces
used to data size and
sample the strengthens
encrypted encryption.
image,
adding
another layer
of security.

xv Improved TANVEER With the rise The authors The scalability

Twofish UL HAQ, of the of the algorithm
i introduce
Algorithm: A TARIQ Internet of for larger or
Digital Image SHAH, Things (IoT) substitution high-resolution
Enciphering GHAZANF and images remains
boxes (S-
Application AR increasing uncertain, as


FAROOQ cyber threats, boxes) testing focused

SIDDIQUI, secure on standard
derived from
MUHAMM communicati image sizes. Its
AD on has a unique dependence on
ZAFAR become a top specific
mathematical
IQBAL, priority. This hardware or
IBRAHIM paper structure computational
A. focuses on resources for
called the
HAMEED, enhancing optimal
(Senior the multiplicativ performance is
Member, TWOFISH not extensively
e group of
IEEE), encryption addressed.
AND algorithm by units in a Additionally,
improving its the use of
HUMA chain ring.
mathematical specialized
JAMIL complexity. These S- chain ring-
The authors based S-boxes
2021. boxes
introduce may pose
substitution increase the compatibility
boxes (S- challenges with
algorithm’s
boxes) systems not
derived from algebraic designed for
a specific such structures.
complexity,
mathematical
structure, enhancing
which adds
security. The
significant
algebraic improved
complexity
algorithm
to the
encryption processes
process.
64-bit data
Additionally,
the blocks and
algorithm’s
 

efficiency is uses fewer

improved by
subkeys,
processing
64-bit blocks which
throughout
simplifies
and reducing
the number operations
of subkeys
and
required.
improves
computation
al efficiency.
xv Fast and Secure Rasha S. The rapid The The limitations
Image Ali, Maha evolution of methodology include a
ii
Encryption Khalil network combines the dependency on
System Ibrahim, technologies, Trivium the chaotic
Using New Saad N. where algorithm map’s
Alsaad computer and chaotic randomness,
Lightweight
2024. vision and Henon maps. which could
Encryption the Internet Trivium is affect
seamlessly used in a performance
Algorithm
merge, poses modified under certain
significant block-wise conditions, and
challenges in manner a focus on
safeguarding rather than 256x256 image
end-user bit-wise, sizes,
data. To while the potentially
address the Henon map limiting its
growing generates scalability.
need for random Additionally,
secure data inputs for while the
in smart internal algorithm
environment initialization shows robust


s, and key statistical

lightweight permutation. performance, its
and efficient This effectiveness
security dynamic and under advanced
solutions are adaptive attacks or in
in high approach diverse real-
demand. ensures world
Notably, the higher applications
Trivium, a security and needs further
well-known better exploration
encrypt resource
stream efficiency
cipher
(eSTREAM)
project, has
been the
target
of several
attacks. This
paper
proposes an
innovative
approach for
fast and
highly
secured
image
encryption
using byte
scrambling
and a
modified
version of
the
Trivium


algorithm.
The Henon
map is
utilized to
generate
random
numbers for
permutation,
reducing
time and
increasing
security.
Performance
is measured
on 100
colour
images with
different
several tests.
The
results
present a fast
and robust
security
system for
the
transmission
process. It is
more
efficient than
the
traditional
Trivium
method. The
encrypted-
 

decryption
time is
reduced by
1:24 times,
making it a
quick
security
system for
surgical
telepresence
and
multimedia
visuals.
xv A Wentao This paper The While the
proposed approach is
ii multidimension Kang 2024. investigates
system innovative,
al image various applies potential
multidimensi challenges
encryption and current
onal chaotic include
decryption information maps for computational
both overhead due to
technology encryption
encryption the high
technologies and dimensionality
decryption. of chaotic maps,
that are
These maps which might
growing enhance the limit its
randomness practicality for
quickly,
and real-time
especially complexity applications.
of the Additionally,
different
encryption like most chaos-
algorithms process, based
which makes encryption
for
the system methods, the
encrypting robust algorithm could


and against be vulnerable if

common the key
decrypting
attacks. The generation
unique algorithm mechanism or
uses chaotic
patterns in
compression parameters are
image data. and not managed
transformatio securely.
The fast
n techniques
Fourier to convert
image data
transform
into highly
(FFT) has secure
formats,
emerged as a
incorporating
powerful hyperchaotic
sequences to
tool with the
generate
potential to encryption
keys. The
significantly
encryption
impact workflow is
optimized
encryption
for high-
technology. dimensional
data,
Current
supporting
encryption larger and
more
technology is
complex
designed to image
datasets.
encrypt and
decrypt


images and
is easy to
decipher.
Hence, this
paper
presents the
proposed
encryption
and
decryption
algorithm to
improve
encryption
technology
based on an
improved
Euler
integral and
FFT; in
particular,
the algorithm
possesses a
more


complex and
multidimensi
onal way to
improve
encryption
technology.
The
improved
algorithm
offers the
ability to
efficiently
process and
analyze vast
amounts of
image data.
The
proposed
approach of
matching the
proposed
encryption
and
 

decryption
algorithms
can be
applied to
cloud
computing
and quantum
computing
technology.
xi A NEW Amer With the rise The study the method’s
IMAGE Jelwy, of proposes a complexity can
x
ENCRYPTION Muthana S. multimedia new image lead to high
AND Mahdi and the encryption computational
DECRYPTION 2023. Internet, and demands,
ALGORITHM protecting decryption limiting its
BASED ON A sensitive algorithm applicability in
FOUR data, that relies on real-time or
especially on a four- resource-
DIMENSIONA
IoT devices, dimensional constrained
L DYNAMIC is crucial. hyper- environments,
This study chaotic such as IoT
SYSTEM
proposes a system to devices. The
new image enhance encryption
encryption multimedia process’s
and security. It sensitivity to
decryption uses the precise key
algorithm unpredictabil management is
that ity of chaotic another
leverages a systems to concern, as
Latin Square ensure robust slight key


matrix and a encryption. deviations can

four- A Latin cause
dimensional square decryption
hyper- matrix is failure. While it
chaotic integrated to demonstrates
system to add strong
enhance complexity, performance
security. The enabling metrics, the
chaotic effective algorithm may
sequences diffusion and face practical
generated by permutation constraints
this system of image when
are used to pixels. implemented on
permute and Statistical devices with
diffuse evaluations limited
images of such as processing
any size, histograms, power and
producing correlation memory.
encrypted coefficients,
images that and
are highly sensitivity
secure. tests confirm
Key the
performance algorithm’s
metrics, such high
as encryption
histograms, quality, with
correlation NPCR at
coefficients, 99.68% and
and key UACI at
sensitivity, 33.47%. Its
were large key
analyzed to space (10¹⁶⁸)
evaluate the provides
algorithm. resistance


The results against

show that the brute-force
encrypted and
images have statistical
a uniform attacks.
histogram,
entropy
values close
to 8 (ideal
for
randomness),
and strong
resistance to
attacks.
Specifically,
the algorithm
achieved a
99.68%
NPCR
(Number of
Pixels
Change
Rate) and a
UACI
(Unified
Average
Changing
Intensity) of
33.48%,
indicating
robustness
against
statistical,
predictive,
differential,
 

and brute-
force attacks.
Its vast key
space and
sensitivity to
key changes
make it
highly secure
for modern
applications.

xx Multiple-Image Lee Mariel This abstract The While the

Fusion introduces a algorithm method
Heucheun
Encryption novel begins with demonstrates
(MIFE) Using Yepdia, multiple- the spectral strong
Discrete Cosine image fusion of encryption
Alain
Transformation encryption multiple performance,
(DCT) and Tiedeu and algorithm watermarked including
Pseudo that images using robustness
Zied
Random integrates Discrete against brute
Lachiri spectral Cosine force, statistical,
Number
fusion of Transformati and differential
2020.
Generators watermarked on (DCT) attacks, its
images with and low-pass limitations
advanced filtering to include
chaotic create two computational
generators to multiplexed overhead
address the images. The associated with
shortcomings resulting the chaotic
of traditional images are systems and
1D chaotic broken into spectral
maps. The smaller transformations.
algorithm blocks, Additionally,
uses three which are the process


chaotic scrambled might face

systems— based on challenges in
Logistic- sequences real-time
May (LM), generated by applications due
May- chaotic to the
Gaussian systems complexity of
(MG), and (Logistic- operations like
Gaussian- May, May- DCT-based
Gompertz Gaussian, spectral fusion
(GG)—to and and chaotic
enhance Gaussian- scramblin
unpredictabil Gompertz).
ity and The fusion of
encryption these
strength. The scrambled
process blocks is
begins by then
applying performed
discrete using a
cosine nonlinear
transformatio mathematical
n (DCT) and formula
a low-pass derived from
filter to Cramer’s
merge rule to create
watermarked encrypted
images into outputs. The
two decryption
multiplex process
images in the reverses
spectral these steps to
domain. retrieve both
These the original
images are image and
then divided embedded


into small data without

blocks, any loss.
which are
scrambled
using a
sequence
derived from
the LM
chaotic
system.
Finally, the
scrambled
images are
fused using a
nonlinear
expression
based on
Cramer’s
rule to
generate two
hybrid
encrypted
images.

The
decryption
process
allows for
the retrieval
of hidden
messages
from
watermarked
images
without loss.


Security
analysis and
experimental
simulations
highlight the
algorithm’s
strong
encryption
performance,
supporting
its ability to
secure
multiple
image types,
including
text, with
low Mean
Square Error
(MSE) after
decryption.
This method
ensures
efficient
encryption
and
maintains
high
security.
