Proceedings of 2024 International Conference on Cognitive Robotics and Intelligent Systems (ICC - ROBINS 2024)

IEEE Xplore Part Number: CFP24UD1-ART; ISBN: 979-8-3503-7274-8

Improving QR Code Security using Multiple

Encryption Layers
2024 International Conference on Cognitive Robotics and Intelligent Systems (ICC - ROBINS) | 979-8-3503-7274-8/24/$31.00 ©2024 IEEE | DOI: 10.1109/ICC-ROBINS60238.2024.10533884

S. Dhivyalakshmi Narayanan Dr. S. Prabhu E. Padma

PG Scholar, “Department of Computer Associate Professor, Department of Assistant Professor, Department of
Science and Engineering Computer Science and Engineering Computer Science and Engineering
Nandha Engineering College(Autonomous) Nandha Engineering College Nandha Engineering College
Erode, Tamil Nadu, India (Autonomous) (Autonomous)
[email protected] Erode, Tamil Nadu, India Erode, Tamil Nadu, India
[email protected] [email protected]

Abstract— Growing concerns about identity theft and privacy are counterfeiting is possible, particularly when it comes to
brought on by the increased sharing of digital information, which merchant account representation. This can result in serious
leaves data open to quick changes while in transit. Digital data financial losses and problems with payment security. Severe
must be protected from hackers. This study presents a novel risks can also arise from malicious links that hackers encode in
approach to improve QR code security and fight counterfeiting. QR codes [6, 7].
This technique strengthens QR code anti-counterfeit measures by
integrating the AES algorithm with Rubik's Cube Principle. The II. LITERATURE SURVEY
Rubik's Cubic Principle is used to separate the RGB channels in
the QR code image. Following the encryption of these channels The effective solution to enhance the security of digital data
using the Advanced Encryption Standard (AES) encryption key, is to use QR code-based security authentication for payments.
the encrypted image is combined with image steganography and is Secret-sharing is accomplished through VSS (Visual Secret
subjected to linear feedback shift register (LFSR) processing. Test Sharing), which is also known as VC (Visual Cryptography)
results confirm the efficiency of the suggested plan, exhibiting [8]. Using this method, an original image is encoded into n
strong encryption and flawless covert abilities. noise, like shadows or shares, so that no individual share will
Keywords- QR code security, AES algorithm, Rubik’s Cube
display. Recovery of the original secret is only possible by
Principle, image steganography, LFSR.
merging the qualifying shares [9]. When it comes to QR code
I. INTRODUCTION payment authentication, the fundamental idea is to use a (2; 2)
scheme of VSSQR. Two QR code shares, designated SC1 and
The ease of mobile payments is just one of the many features
SC2, comprise the original secret binary image. Details about a
that mobile phones offer in this age of rapid technological
merchant are contained in both shares. While SC2 is kept on the
advancement. Due to its ease of use and speed in financial
server and the merchant's desk, SC1 is shown on the wall or
transactions, this method has become increasingly popular [1].
desk of the merchant. With any mobile device, customers can
Applications that take advantage of this technological shift are
access the server by scanning both SC1 and SC2. Decoding the
emerging as internet payment solutions become more
information in SC1 and determining whether the stacking result
widespread across industries and payment processes become
of SC1 and SC2 matches the secret image further confirms the
digital. Credit cards and NFC-based transactions are just two
authenticity of SC1 [10] are the first few steps in the
examples of the many ways that online payments could evolve
preliminary authentication process. The drawback of this
in the future. However, as technology advances, so does the risk
technique is that when multiple shares are overlaid, the VSS
of security breaches. Studies on Online Payment System
system can retrieve a hidden image without requiring complex
Implementations show how important safety is to both
computations because it makes use of the Human Visual
businesses and consumers. The availability, confidentiality, and
System (HVS).
integrity of the system are put at risk by payment transactions,
which are vulnerable to theft, fraud, and counterfeiting. Internet III. PROPOSED WORK
payment systems need to give security measures top priority in
order to overcome these security issues. The ability of QR A. Rubik’s Cube Principle
codes, or two-dimensional barcode matrices, to quickly encode • Foundation Establishment: The pixels in the image are
and store large amounts of data makes them a practical solution utilized as the basis for the Rubik's cube algorithm.
[1,2]. The speed and simplicity of QR codes make them useful
• Jumbling Process: This process is used to modify the
in important fields like healthcare, education, and finance [3, pixel's value and shift columns or rows by using modulo
4]. These 2D barcodes are advantageous because they can 2 to jumble the original image.
easily be retrieved by QR code readers and have high
information density and robustness. They can also correct • XNOR Operation: The XNOR operation is applied for
errors. The speed at which QR codes are being adopted for even columns with Kr and even rows with Kc by
payments, security authorization, and other uses highlights their flipping the keys Kc and Kr.
effectiveness. Though QR code technology is open-source,

979-8-3503-7274-8/24/$31.00 ©2024 IEEE 845

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on May 31,2024 at 06:13:11 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of 2024 International Conference on Cognitive Robotics and Intelligent Systems (ICC - ROBINS 2024)
IEEE Xplore Part Number: CFP24UD1-ART; ISBN: 979-8-3503-7274-8

• Row and Column Operations: The row operations are • Encryption Preparation: IN this process, a linear
executed using Kr and column operations are executed feedback shift register (LFSR) is used to generate
using Kc. random numbers for encryption and decryption.
• Decryption: This process id used to decrypt the • Initial Encryption: The encryption process is initialized
encrypted image by using a specific number of iterations by loading a color image and use its initial values to
and keys. initialize LFSRs.
• Component Identification: This identification process • Secondary Encryption: After initial step, the new
has used the single-tone component as the foundation of random numbers will be generated by modifying LFSR
the Rubik's Cube method to process black-and-white or initial values with specific numbers based on the image's
grayscale images. dimensions.
• Challenge Recognition: This recognition acknowledges • Decryption Initialization: The decryption phase will be
the challenge posed by RGB images, where each pixel initialized by loading the encrypted color image with the
contains multiple tone components. same initial values.
• Individual Application: Finally, the Rubik's cube • Decryption Process: This consecutive step supplement
method is applied individually to each element of the LFSRs with specific numbers to recreate the same
RGB image [11]. random numbers used for column and row decryption.
[16,17].
B. AES Algorithm
The AES algorithm encrypts the random key of the Rubik’s Fig. 2. illustrates how the of Linear feedback shift
cube principle, producing an encrypted random key. The register and Image steganography function.
process is:
• Bytes Substitution: This process substitutes the bytes in
the block text based on predefined S-boxes to enhance
complexity.
• Move the rows: All the rows will be replaced except the
1st through a permutation step to introduce further
variation. Fig. 2. Working of Linear feedback shift register and Image steganography
• Mix the columns: A Hill cipher approach is usedto mix
the columns by combining the blocks of columns to
increase complexity. IV. METHODOLOGY

• Add the round key: This process XOR the message with The following sections contain two modules:
the corresponding round key in the final step to ensure • Encryption Module
the secure encryption over multiple iterations. • Decryption Module
[12,13,14].
Fig. 1. illustrates how the AES algorithm and the A. Encryption Module
Rubik’s cube principle function. The process starts with a QR code image. The image is split
into red, green, and blue parts. Then, a random key encrypts
these parts using AES. These encrypted parts are combined into
one encrypted image. Next, this image is further encrypted with
the cover image and a linear feedback shift register. Finally, the
new encrypted image is hidden within the cover image, creating
a new image that hides the encrypted QR code. Fig. 3. shows
block diagram for encryption process.
Fig. 1. Working of Rubik’s cube principle with AES (Advanced
Standard Encryption)

C. Image Steganography and Linear feedback shift register

• Steganography Initialization: The process is initialized
by concealing data within an image file, known as image
steganography.
• Image Terminology: The cover image will be defined as Fig. 3. Block diagram
the original image and the stego image as the result after
steganography [15].

979-8-3503-7274-8/24/$31.00 ©2024 IEEE 846

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on May 31,2024 at 06:13:11 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of 2024 International Conference on Cognitive Robotics and Intelligent Systems (ICC - ROBINS 2024)
IEEE Xplore Part Number: CFP24UD1-ART; ISBN: 979-8-3503-7274-8

B. Decryption Module C. Peak Signal-To-Noise Ratio

Decryption works in the opposite way of encryption. First, The ratio of an image's maximum potential power to the
the image is split to get the cover image and the new encrypted corrupting noise power that degrades the image's representation
image. Then, the new encrypted image is decrypted using the quality is called the PSNR (Peak Signal-To-Noise Ratio). An
cover image and XOR. Next, the decrypted picture is split into image must be compared to the best possible clean image with
three parts, each receiving the encrypted random key. Finally, the highest power to find out its PSNR. PSNR is defined as
the three decrypted parts are combined to rebuild the QR code follows (2):
image. (𝐿−1)2 𝐿−1
𝑃𝑆𝑁𝑅 = 10 log10 ( ) = 20 log10 ( ) (2)
𝑀𝑆𝐸 𝑅𝑀𝑆𝐸
V. RESULT Here, L is the number of maximal intensity levels that
can exist in an image, with 0 being the minimum intensity level
A. Histogram Analysis that is assumed. MSE (Mean Squared Error) & it is defined as
The main purpose of using histograms is to analyze the pixel (3):
distribution in an image and extract relevant information [18]. 𝑀𝑆𝐸 =
1
∑𝑚−1 ∑𝑛−1 2
(3)
𝑚𝑛 𝑖=0 𝑗=0 (𝑂(𝑖, 𝑗) − 𝐷(𝑖, 𝑗))
By minimizing pixel variations, this encryption technique
prevents attackers from obtaining personally identifiable where O signifies the original image's matrix data. D is the
information about the image. The histograms for the original degraded image's matrix data. The variables m and i denote the
QR code image and the encrypted QR image are shown in Fig. number of rows” of pixels and the row index, respectively, in
4. The histogram of an image displays the frequency of various the image. The variables n and j denote the number of pixels in
R, G, and B levels in the image relative to each other. For every each column and the image's corresponding column index,
R, G, and B level, pixel counts are plotted on the y-axis and respectively. The root mean square error is known as RMSE.
intensity values are plotted on the x-axis. The encrypted image's
standard histogram outcome makes it more difficult for
unauthorized parties to extract any information and makes it TABLE I. COMPARISON
more difficult for them to recreate the original image. QR code Encrypted NPCR PSNR
Image

99.6500 27.7860
VSS

Proposed 99.9058 27.9258

Multiple
Encryption

99.5946 27.6421
VSS

Proposed 99.8921 27.8861

Multiple
Fig. 4. Histogram of Original and Encrypted Image
Encryption

B. Number of Pixel Rate Change test Table 1. shows the comparison of VSS and proposed
multiple encrypted image.
One of the statistical tests used to assess how resilient an
encrypted image is to statistical attacks is the NPCR ("Number VI. CONCLUSION
of Pixel Change Rate"). NPCR is used to quantify the rate of The suggested method has shown increased efficiency and
change between the original and encrypted image pixels [18]. effectiveness, as seen in the results obtained from experiments
The mathematical formula (1) is used in the calculation of with a group of images. These results allow for comparison with
NPCR values. other methods. Tests conducted show a high level of security
Σ𝑁 𝐷(𝑖,𝑗)
𝑁𝑃𝐶𝑅 = ∑𝑀 𝑖=1
𝑗=1
∗ 100 (1) achieved by the multi-encryption algorithm, with a score
𝑀∗𝑁
M and N stand for the interferogram's width and height, of 99.94%. This indicates its ability to resist different types of
respectively, in the encrypted form. C1(i,j), the interferogram, attacks. Furthermore, the results suggest that attackers cannot
is encrypted before a pixel change.C2(i,j) is the encrypted extract information about the original image data from the
interferogram after a pixel change. D(i,j) is a bipolar network. resulting images.

979-8-3503-7274-8/24/$31.00 ©2024 IEEE 847

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on May 31,2024 at 06:13:11 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of 2024 International Conference on Cognitive Robotics and Intelligent Systems (ICC - ROBINS 2024)
IEEE Xplore Part Number: CFP24UD1-ART; ISBN: 979-8-3503-7274-8

REFERENCES [10] SongWan, Guozheng Yang, Lanlan Qi, Longlong Li, Xuehu Yan_and
Yuliang Lu, “Multiple security anti-counterfeit applications to QR code
[1] P. De, K. Dey, V. Mankar, “An assessment of qr code as a user interface payment based on visual secret sharing and QR code”, Mathematical
enabler for mobile payment apps on smartphones”, IndiaHCI’15, Biosciences and Engineering Volume 16, Issue 6, 6367–6385.
December 17-19, 2015.
[11] Srushti Gandhi, Taral Butani, Ravi Gor, “Image Encryption System using
[2] Ivan Jelic and Dina Vrkic, “QR codes in the library - Does anyone use Rubik’s Cube Algorithm”, IOSR Journal of Computer Engineering
them?”,2013. (IOSR-JCE), Volume 24, Issue 4, Jul. –Aug. 2022.
[3] S. Agrawal, "Impact of Error Filters on Shares in Halftone Visual [12] https://www.geeksforgeeks.org/advanced-encryption-standard-aes/
Cryptography", International Conference on Computer Science,
Engineering and Applications, 2012. [13] https://www.geeksforgeeks.org/block-cipher-modes-of-operation/
[14] Ako Muhamad Abdullah, “Advanced Encryption Standard (AES)
[4] Z. Zhou, G. R. Arce, and G. D. Crescenzo, "Halftone visual
Algorithm to Encrypt and Decrypt Data”, Cryptography and Network
cryptography", Image Processing, 2003 International Conference on
Security, June 16, 2017.
IEEE, I-521-4.vol.1, 2003.
[5] Kamal, Sawsan, Ameen, Basheer, “A New Method for Ciphering a [15] Nandhini Subramanian , Omar Elharrouss , Somaya Al-Maadeed, And
Message Using QR Code”, Computer Systems Science and Engineering, Ahmed Bouridane, “Image Steganography: A Review of the Recent
2016. Advances”, IEEE Access, 2017.
[6] P. Kieseberg, M. Leithner, M. Mulazzani, et al., Qr code security, [16] Krishnapriya P V, Smitha Suresh, “Image Security Using Linear
Proceedings of the 8th International Conference on Advances in Mobile Feedback Shift Register”,International Journal of Innovative Science and
Computing and Multimedia, 2010. Research Technology, Volume 2, Issue 6, June – 2017.
[7] Katharina Krombholz, Peter Fruhwirt, Peter Kieseberg, Ioannis Kapsalis, [17] Salah T. Allawi, Dr.Jamila H. Al-A’meri , “Image yption Based on Linear
Markus Huber, and Edgar Weippl, “QR code security: A survey of attacks Feedback Shift Register Method”, Al-Sadeq International Conference on
and challenges for usable security”, International Conference on Human Multidisciplinary in IT and Communication Science and Applications,
Aspects of Information Security, Privacy, and Trust, 2014. 2016.
[8] Z. Wang, G. R. Arce, and G. Di Crescenzo, “Halftone visual cryptography [18] Salah Taha Allawi, Nada Abdul Aziz Mustafa, “Image encryption based
via error diffusion”, IEEE Trans. Inf. Forensics Security, 4, 383–396, on combined between linear feedback shift registers and 3D chaotic
2009. maps”, Indonesian Journal of Electrical Engineering and Computer
Science, Vol. 30, No. 3, June 2023,
[9] A. Beimel, “Secret-sharing schemes: A survey”, International Conference
on Coding and Cryptology, 2011.

979-8-3503-7274-8/24/$31.00 ©2024 IEEE 848

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on May 31,2024 at 06:13:11 UTC from IEEE Xplore. Restrictions apply.