An Enhanced Triple Prime Encryption Approach

2023 First International Conference on Advances in Electrical, Electronics and Computational Intelligence (ICAEECI) | 979-8-3503-4279-6/23/$31.00 ©2023 IEEE | DOI: 10.1109/ICAEECI58247.2023.10370817

for Image Encryption in LSB Steganography

B. Bharathkumar a, S. Logeswar a, S. Sudharsun a, B. Karthikeyan a*
a
School of Computing, SASTRA Deemed University, Thanjavur, India.

*Corresponding author - [email protected]

Abstract— The internet is developing rapidly, which leads to a In general, the term cryptography is an essential tool to
sharp increase in security levels. Compression of data is used to protect sensitive information in the digital era. There are lots of
save memory securely. At the same time, the protection of data is cryptographic algorithms are there to protect the data,
very important. The effective compression method gives efficient developers can choose one of the main algorithms depending
and easy-to-connect data. To protect data securely cryptographic on security purposes. Cryptography is a branch of cryptology
and steganographic algorithms play a crucial role in the modern where data change in the form of readable format to unreadable
world. There are a lot of algorithms in cryptography to protect format. It is classified as classical and modern cryptography.
the data. In this study, three different ways to protect the data Classical cryptography defines substitution and transposition
were proposed. Image scrambling is also an efficient way to
techniques. Modern cryptography specifies symmetric and
protect the data. Discrete Cosine Transform (DCT) and Arnold’s
cat map techniques were used to compress and encrypt the
asymmetric key cryptography. In the first type, the size of
image. Multiple encryptions were used to encrypt the plain text. chipper text is small or the same as the original text and it uses
The first level is an alphanumeric scrambling sequence and in the a single key for encryption as well as decryption. It needs a
second level, a new algorithm named Triple Prime Encryption large amount of data to transfer and took less time. It provides
(TPE) gives good results in security lastly one of the public key confidentiality without integrity and authenticity. In the second
cryptography algorithms RSA was used to encrypt the text. The type two keys are used for encryption and decryption and the
combination of Lempel Ziv Welch (LZW) and Huffman coding size of the chiper text is larger than the original text.
was used to compress the lossless data which provides an efficient
Steganography is the way that hides the data inside the
security level and high compression ratio. Finally, the encrypted
data is encoded into a scrambled image. To analyze the
image, audio, video, text file, etc. In ancient times
performance of this compression some parameters were steganographic techniques are mostly done physically. The
performed which is compression ratio, compression time, savings combination of both cryptography and steganography gives
percentage, Bits per pixel, and histogram analysis. To analyze the more benefit to the sender as well as the receiver. First, encrypt
performance of stego images Mean Squared Error (MSE), Peak the plain text using any one of the cryptographic algorithms
Signal to Noise Ratio (PSNR), and Structural Similarity Index and then hide the encrypted text in any format files.
(SSIM) were performed. Experimental results show the best
Compression is a technique that is used to diminish the size
accuracy of the performance of the analysis.
of files and the security of the data will be increased while
Keywords— Discrete cosine transform, Arnold’s cat map, LZW, those compressions are used in cryptography and
Huffman coding, RSA. steganography. Compression is classified into two types which
are lossy and lossless compression. Lossy compression is used
to compress the image, audio, and video with some data loss.
I. INTRODUCTION Lossless compression is used to compress text and programs
With the ravishing development in digital technologies without any loss of data.
security plays a vital role while transferring files which is
related to Military and Defense. A chaotic map is one of the II. LITERATURE SURVEY
vital roles which is a very interesting technique used to protect
the data securely. Chaotic maps are mathematical functions In this work, the image scrambling method was proposed
that exhibit complex and unpredictable behavior over time. In using some novel encryptions. In the first step the image is
recent years, image encryption algorithms have expanded in scrambled using the Hilbert curve and a novel method bit-level
the field of defense to prevent data from attackers. The goal of permutation using cyclic shift and double cyclic shift was used
image encryption is to protect the data from unauthorized to encrypt the image [1]. The main disadvantage of this process
individuals. Image scrambling, also known as image is only for grayscale images and the advantage is correlation
obfuscation, is a technique where the pixel values are values are decreased. Initially performs Arnold transform to
scrambled over the original image and the data is in the form of change pixel values and performs EX-OR operation for each
unreadable and protected from attackers. This is done by a pixel value. The results clearly explain about the attackers can
mathematical algorithm that rearranges the values of pixels in a easily identify the message in histogram analysis. It does not
randomized way. It is often used for security purposes, such as clear about the reversible process and robustness [2]. Initially,
protecting sensitive images and sensitive data. Another way to the image is scrambled using pixel scrambling operation with
protect data is cryptography and steganography. circular shift and swapping operation up to n times of round.

Authorized licensed use limited to: J.C. Bose University of Science and Technology YMCA Faridabad. Downloaded on November 06,2024 at 10:28:05 UTC from IEEE Xplore. Restrictions apply.
 The high robustness and quality of the image improved [3]. technique is applied for the image which gives a good pixel
Firstly, the image is scrambled using Arnold’s cat map and value before the compression called pre-processing. It gives
encrypted using a famous symmetric algorithm Advanced Data high capacity and high compression efficiency [14]. People use
Encryption (AES). It uses a Butterworth method of high-pass multimedia tools for sharing files. Sometimes those files are
filter to enhance the security level of medical images from the larger in size and File compression is one of the methods to
loss of data at encryption and decryption. Three bits of reduce the size of the files. Here the text file is compressed
Arnold’s cat map are going to AES. It is used for pre- using LZW compression which is mainly used to reduce larger
processing phase to gain more security. A large number of datasets to reduce the space requirements without affecting
parameters can reduce the time for image encryption and other parameters [15]. The volume of medical data is high
decryption [4]. This paper proposes an image that is scrambled under iot domain so they use compression techniques. For
by Arnold’s cat map and Fast Discrete Curvelet Transforms some techniques, the quality of the image is reduced. Here the
(FDCUT) and compressed using Discrete Cosine Transform combination of Huffman and LZW techniques is proposed
(DCT) and Singular Value Decomposition (SVD) conversion where binary information reaches information mapping. If the
and performs all those inverse processes in the medical image map gets simplified, it gives completely a different piece of
in Electronic Medical Records (EMR). It gives high robustness information [16]. Here the data is encrypted using the AES
and key uniqueness [5]. encryption technique and compressed using LZW lossless
compression method. It reduces the storage space of backup
The plain image is converted from RGB to YCbCr and data and encrypts the data in blocks. If the encryption key is
compressed using DCT. Zigzag and vertical methods are used leaked it does not affect the data because of the combination of
to read the coefficients. The lowest values of indexed vectors compression and encryption [17]. In this work, lossy and
maximize the best scan. The advantage of this process is lossless compression were performed. Encryption is achieved
simple and high performance [6]. Scrambling techniques using the famous RSA algorithm and lossless compression is
always confuse attackers by scrambling the values of images or also performed after the process of encryption. The image is
data. This paper uses chaotic signals and chaos keys for compressed using DWT compression and uses LSB
encryption and produces good results. It gives high capacity steganography [18]. Steganography is a method of transmitting
and finding the frequency of audio is difficult [7]. Three maps data through images, photos, video, and audio. Here the data is
sine, tent, and logistics in chaotic system handled. The first two encrypted using the RSA algorithm and hidden inside an image
maps are used in dynamic row scrambling and the logistic [19]. Three techniques were proposed in this paper. Plain text
system is used in zigzag scrambling and also this process is directly encrypted using RSA asymmetric algorithm. Image
provides high robustness [8]. LZW and the optimality of LZW steganography and pseudo numbers to choose a random
were proposed. Checking whether the dictionary is full or not. number of pixels value after encrypting the image.
If it is full, it deleted the last recently used dictionary phrases. Performances were analyzed using MSE, PSNR, and
Another method is proposed if no one can delete the string that Histogram analysis [20] [23]. In this paper, multiple
is added to the dictionary and it also checks the previously encryptions were proposed as an innovative approach. Odd
matched string to retrieve it. It gives a high compression ratio lines are encrypted as four levels of encryption and even lines
compared to other methods [9]. A combination of Huffman are encrypted as a simple scrambling method. Later the
coding and LZW is performed. Firstly, the text is compressed encrypted text is embedded into the image. For performance
using Huffman coding a lossless compression algorithm and analysis, MSE and PSNR metrics were used. An effective
the resultant Huffman coding is compressed using LZW. It is a algorithm is used for multiple encryptions and the security
cost-effective method and gives good results in savings level is high [21].
percentage and compression ratio [10]. In this work, the image
is compressed using column dimensions and the result of
prediction error in this process is going to be compressed using III. PROPOSED METHODOLOGY
the lossless compression algorithm Huffman coding which 3.1 Alpha-numeric Scrambling Sequence
results in average code length the used parameters to analyze It is the first level of encryption where the encryption is
the performances are compression ratio, PSNR and normalized achieved by scrambling the alphabet of the plain text. Fig 1
correlation [11]. Yoruba is a language spoken by people in shows an alphanumeric scrambling table.
Nigeria. Here the process is to compress the textile using DWT
and Huffman which is in the Yoruba language. Initially, the
Yoruba text file is converted into signals and those signals are
compressed using DWT and the resultant values will be
compressed using Huffman compression [12]. Digital Imaging
and Communications in Medicine (DICOM) images have
played a vital role in our day-to-day life [22]. In this work, Fig 1 Scrambling alpha-numeric table
linear predictive coding and Huffman coding are used to
compress the images. Histogram analysis was used to analyze 3.2 Triple Prime Encryption (TPE)
the performances of the images [13]. Reversible data hiding is Triple prime encryption is a new methodology that is either
one of the steganographic techniques which is applied when symmetric or asymmetric key cryptography. It contains only
there is an image compressed by lossless image compression. It one key to encrypt and decrypt. Each alphabet consists of one
hides the secret image and transfers over transmission without triple digit prime number and there is a formula for encryption
any doubt to intruders. A modified LZW compression and decryption.

Authorized licensed use limited to: J.C. Bose University of Science and Technology YMCA Faridabad. Downloaded on November 06,2024 at 10:28:05 UTC from IEEE Xplore. Restrictions apply.
 For encryption, the formula is CT = 2e M LZW is a dictionary-based compression and Huffman is
Where e is a key which is a prime number, CT is a chiper text assigning some shorter code to the frequently repeated data.
and M is a plain text. Generally, the LZW compression alone gives good results and
For decryption, the formula is M=CT/2e here the combo of both algorithms provides more security and
3.3 RSA Methodology is more efficient.
The final level of encryption is RSA public key 3.5 Image Encryption
cryptography which uses two keys for encryption and The encryption process of the plain text is done parallelly.
decryption. Fig 2 shows the working process of the RSA On the other hand, the plain image is going to be encrypted
algorithm. using some other techniques. First, the plain image is
compressed using Discrete Cosine Transform and the resultant
image is scrambled using Arnold’s Cat Map which is one of
the scrambling techniques and chaotic map theories that a
small change in the input there is a huge impact on the resultant
output.
3.6 LSB Steganography
In common the word steganography is defined as hiding a
secret message inside the image, audio, video, or textiles. Here
image steganography is proposed. And the method to hide the
message is Least Significant Bit. It is a common method in
steganography where the pixels of an image and binary values
of a plain text is interchanged using the Least Significant Bit
methodology. Fig 4 shows the working process of the proposed
methodology.

Fig 2 RSA Workflow

In this work, φ(n) described as l. The mathematical values

of RSA are mentioned below. Fig 3 shows the mathematical
values of RSA algorithm which is used in this paper.

Fig 3 Mathematical values of the RSA algorithm

Fig 4 Workflow

3.4 Lossless Compression

Lempel Ziv Welch compression is a lossless compression IV. RESULTS AND DISCUSSION
which is compressing the data without any loss. Huffman In conclusion, the encrypted chiper text is embedded into a
coding is also a lossless compression algorithm. In previous scrambled image using LSB steganography. To ensure the
work, only one compression is performed which gives the performances of the compressed image and stego image are
analyzed by the compression ratio, compression time, savings
results in a good way. Here the proposed methodology is a
percentage, Bits per pixel, Statistical analysis, MSE, and
combination of LZW and Huffman coding which gives a high PSNR.
compression ratio and robustness.
4.1 Compression Ratio
Firstly, the encrypted message using the final level of
It is calculated between the ratio of the normal image and
encryption is going to compress in the next level of this compressed image with the definition of high compression
process. Initially, LZW compresses the encrypted message and ratio gives high quality of the compressed image.
then the resultant is compressed using Huffman coding.

Authorized licensed use limited to: J.C. Bose University of Science and Technology YMCA Faridabad. Downloaded on November 06,2024 at 10:28:05 UTC from IEEE Xplore. Restrictions apply.
 compression ratio = size of the plain image/size of the
compressed image

4.2 Compression Time

This parameter describes the time consumed for the
compression of images and rebuilds the original image. The
lowest value is considered a good performance level.
4.3 Savings Percentage
This parameter describes the percentage of savings from an
uncompressed image to a compressed image. Here also higher
value is the best accuracy level.
sp = (original file size-compressed file size)/original file size
4.4 Bits Per Pixel
The amount of data required to represent one pixel of a
compressed image is the work of the Bits Per Pixel parameter.
BPP= number of bits in compressed image/ total number
of pixels in an image
4.5 Statistical Analysis
The statistical analysis describes the relationship between
the Plain image and the scrambled image. Two types of
statistical analysis are histogram analysis and correlation
analysis. The resultant values of both analyses are mentioned
below. Fig 5 Input and output images

4.6 Mean Squared Error Fig 6 shows the histogram analysis of the plain image and
The pixel difference between the original image and the encrypted image.
encrypted image.
MSE = (1/n)*∑(Xi-Yi)2
n= number of data points
X= observed values
Y=predicted values

4.7 Peak Signal-to-Noise Ratio

This is a ratio between signal strength and noise that
appears in those signals. Here the value is obtained only the
MSE value is calculated
PSNR=10log10((255)2)/MSE
255 is the maximum gray value for an 8bit representation per
pixel. Table 1 shows the accuracy of the parameters.

Table 1 Accuracy Table

Fig 5 shows the subjective analysis of the images which
represent the output images using lossy compressions.

Fig 6 Histogram analysis

Authorized licensed use limited to: J.C. Bose University of Science and Technology YMCA Faridabad. Downloaded on November 06,2024 at 10:28:05 UTC from IEEE Xplore. Restrictions apply.
 Table 2 shows the correlation analysis of plain images.
Fig 9 shows the graphical representation of the vertical
correlation between plain images and encrypted images.

Table 2 Correlation Values of plain images

Table 3 shows the correlation analysis of encrypted images.

Fig 9 Vertical correlation

Table 3 Correlation Values of encrypted images V. CONCLUSION

The concept proposed in this paper increased the security
Fig 7 shows the graphical representation of the parameters. level of the data to be embedded. It is used for management
who need a very secure means of transport and very less time
to encrypt and decrypt the data This technique will make it
harder for an individual who can access the data of
organizations without authorization. Proposing a new
cryptographic algorithm that is not connected to the general
way or ahead of traditional and also trying the different
chaotic maps in a proper way to enhance the security level.

REFERENCES
[1] Shahna K.U., Anuj Mohamed, “A novel image encryption scheme using
both pixel level and bit level permutation with chaotic map,”Applied
Soft Computing, Volume 90, 2020.
[2] H. Ye, S. Huang and W. Liu, “Research on Image Scrambling Method
Fig 7 Representation of images using lossy compression
Based on Combination of Arnold Transform and Exclusive-or
Operation,” IEEE 4th Information Technology, Networking, Electronic
and Automation Control Conference (ITNEC), Chongqing, China, 2020.
Fig 8 shows the graphical representation of the horizontal [3] V. K. Sharma, R. Jaya Krishna, A. Singh and V. S. Rathore, “Secure
correlation between plain images and encrypted images. Data Hiding Technique using Secret Image Scrambling,” 4th
International Conference on Electronics, Communication and Aerospace
Technology (ICECA), Coimbatore, India, 2020.
[4] M. A. W. Shalaby, M. T. Saleh and H. N. Elmahdy, “Enhanced Arnold’s
Cat Map-AES Encryption Technique for Medical Images,”2nd Novel
Intelligent and Leading Emerging Sciences Conference (NILES), Giza,
Egypt, 2020.
[5] M. F. Mahyudin, L. Novamizanti and S. Sa'idah, “Robust Watermarking
using Arnold and Hybrid Transform in Medical Images,”IEEE
International Conference on Industry 4.0, Artificial Intelligence, and
Communications Technology (IAICT), Bandung, Indonesia, 2021.
[6] A. Messaoudi, “DCT-based compression algorithm using reduced
adaptive block scanning for color image,”19th International Multi-
Conference on Systems, Signals & Devices (SSD), Sétif, Algeria, 2022.
[7] Xingyuan Wang, Xuan Chen, “An image encryption algorithm based on
dynamic row scrambling and Zigzag transformation,”Chaos, Solitons &
Fractals, Volume 147, 2021.
Fig 8 Horizontal correlation

Authorized licensed use limited to: J.C. Bose University of Science and Technology YMCA Faridabad. Downloaded on November 06,2024 at 10:28:05 UTC from IEEE Xplore. Restrictions apply.
 [8] Sattar B. Sadkhab,Ammar M. Raheema,Sinan M. Abdul Sattar, “Design [17] Zeng, T., Wang, Sb, “Research on LZW Algorithm Based on AES in
and Implementation Voice Scrambling Model Based on Hybrid Chaotic Data Backup Under Data Block Compression and Encryption,” Lecture
Signals,”First International Conference of Computer and Applied Notes in Electrical Engineering, vol 827. Springer, 2022.
Sciences (CAS), 2020. [18] Wahab,Osama Fouad Abdel and Khalaf, Ashraf A. M. and Hussein,
[9] Utpal Nandi, Jyotsna Kumar Mandal, “Modified Compression Aziza I. and Hamed, Hesham F. A, “Hiding Data Using Efficient
Techniques Based on Optimality of LZW Code (MOLZW),” Procedia Combination of RSA Cryptography, and Compression Steganography
Technology, Volume 10, 2013. Techniques,” Volume 9,2021.
[10] G. Shrividhiya, K. S. Srujana, S. N. Kashyap and C. Gururaj, “Robust [19] P. B, R. Harish, S. B and V. M, “Image Steganography using RSA
Data Compression Algorithm utilizing LZW Framework based on Algorithm for Secure Communication,” 2021 IEEE International
Huffman Technique,” International Conference on Emerging Smart Conference on Mobile Networks and Wireless Communications
Computing and Informatics (ESCI), Pune, India, 2021. (ICMNWC), Tumkur, Karnataka, India, 2021.
[11] Rahman, M.A., Hamada, M., “A prediction-based lossless image [20] Hatem Nahi Mohaisen, Awad Kadhim Hammoud, “Application of
compression procedure using dimension reduction and Huffman modify RSA cryptography and randomly LSB steganography on color
coding,” Multimedia Tools Appl 82,4081-4105(2023). images of fluid flow in a channel,”International Journal of Nonlinear
[12] Amusa, Kamoli & Adewusi, Adeoluwawale & Erinosho, Tolulope & Analysis and Applications.Volume 12, Issue 2 November 2021.
Salawu, Sule & Odufejo, David, “On the application of wavelet [21] B. Karthikeyan, B. Bharathkumar, G. Manikandan, and R.
transform and Huffman algorithm to Yorùbá language syntax text files Seethalakshmi, “A Combination of RSA Algorithm with Image
compression,” Serbian Journal of Electrical Engineering,2020. Steganography to Ensure Enhanced Encryption,” Second International
[13] Latha. H.R, Rama Prasath, A, “A Hybrid Approach for Lossless Dicom Conference on Electronics and Renewable Systems (ICEARS),
Image Compression Using Combined Approach of Linear Predictive Tuticorin, India, 2023, pp.773-777.
Coding and Huffman Coding with Wavelets,” Communications in [22] Bala Krishnan R., Rajesh Kumar, N., Raajan, N.R., Manikandan, G.,
Computer and Information Science, vol 1697,2020. Srinivasan A, Narasimhan D., "An Approach for Attaining Content
[14] Leelavathi, R., Prasad, M.N.G, “A high-capacity reversible data hiding Confidentiality on Medical Images Through Image Encryption with
scheme in efficient LZW compression,” Multimed Tools Appl 82, 9459– Steganography", Wireless Personal Communication, 127, 979–995,
9476(2023). 2022.
[15] Gupta, S., Yadav, A., Yadav, D. et al., “A scalable approach for index [23] Chhikara, S., Kumar, R., “Information theoretic steganalysis of
compression using wavelet tree and LZW,” Int. j. inf. Tecnol,2022. processed image LSB steganography,”Multimedia Tools Applications,
82 (9), 13595–13615, 2023.
[16] Hossein Mohammadi, Abdulbaghi Ghaderzadeh & Amir Sheikh
Ahmadi, “A Novel Hybrid Medical Data Compression Using Huffman
Coding and LZW in IoT,” IETE Journal of Research, 2022.

Authorized licensed use limited to: J.C. Bose University of Science and Technology YMCA Faridabad. Downloaded on November 06,2024 at 10:28:05 UTC from IEEE Xplore. Restrictions apply.