Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1763

Security Scheme by Using Steganography in

Multi-Model Biometric System
Boby Kumar1; Deepak Sharma2; Dr. Navin Kr. Tyagi3
Department of Computer Science and
Engineering MIT Bulandshahr

Abstract:- The process of automatically identifying a three parameters: matching module, feature matching, and
person based on their physiological and behavioural traits sensing.
is known as biometric recognition. A unimodal or
multimodal biometric system might be based on In unimodal biometric system it is not possible because
physiological or behavioural traits. Some shortcomings in in this systemhaving multiple things and attached with single
the unimodal biometric system include spoofing, non- modalities.
universality, intra-class differences, and noisy data. Using
the face and palm images as templates, we are II. LITERATURE REVIEW
implementing a multimodal biometric system. A person's
palm image can provide a wealth of detailed information S.B. Verma et al. [8] proposed ‘’Comparative Study of
about them. Palm image have three major lines, name as FAST, MSER, and Harris for PalmprintVerification System’’.
principle line, secondary line and wrinkles which are In this paper they have compared the study of feature
unique in nature. Face image have more feature than any extraction of palm print using three different method name as
other biometric traits. Face image is recognized easily in FAST, MSER and HARRIS. In this comparative study they
less time. Combining these two biometric template gives have taken two images of palm names as A and B. After that
higher security. Message image kept secretly in the cover they have applied these three methods, then the FAST feature
image called stego image. In this steganography technique extraction detects 74 points, MSERdetects 128 points, and
we are using palm image as a Message and face image as HARRIS detects 289 points in the hand A image.
a cover image. After getting stego image Identification
process works. The important featureof identification process Qi Xiong et al. [12] proposed ‘’A Modified Chaotic
id De-Stego. This feature of identification will separate the Binary Particle Swarm Optimization Scheme and Its
message image (palm image) from cover image (face Application in Face-Iris Multimodal Biometric
image). After separation of images the decision will take Identification’’ After extracting the feature of both templates
care, weather the image identified or notidentified. the complexity is much higher of extracted feature. To solve
this problem, they have used modified chaotic binary particle
Keywords:- Steganography Technique De-Stego, Face swarm optimization (MCBPSO) algorithm for selecting the
Image, Palm Image, Stego-Image Verification And accurate feature. They have used The Institute of Automation,
Identification. Chinese Academy of Science (CASIA) database for
experiments.
I. INTRODUCTION
S.B. Verma et al. [9] proposed ‘’Performance Analysis of
Biometric Recognition is an act of recognizing the Various fusion method in multimodalbiometric’’ They have
person quickly and automatically based on their physical and analyzed various fusion in different- different process of an
behavioural characteristic. The distinctiveness of these image. They have used four levels of fusion methods name as
features makes biometric recognition systems increasingly sensor level fusion, feature level fusion, score level fusion,
common. Physiological or behavioural qualities cannot be decision level fusion.to process the image. Firstly, they have
lost or stolen because they are unique to everyone. The two applied sensor level fusion in image acquisition. Then they
different types of biometric systems are multimodal and have applied feature level fusionmethod in feature extraction.
unimodal biometric systems. Unimodal biometric systems They have used score level fusion in score matching. And
only recognise single traits, such as only face recognition, finally they have used decision level fusion for making the
fingerprint recognition, iris recognition, etc., whereas decision.
multimodal biometric systems recognise multiple traits, such
as face and palm print recognition, iris and thumb Mustafa Ahmed Shamil et al. [6] proposed “multimodal
recognition, and there can be multiple possible outcomes in biometric system iris and fingerprint recognition based on
this biometric system. Multimodal biometric systems are fusion technique” was published. Basically they have
growing in popularity due to their capacity to inherit researched on multimodal system and they used fingerprint
characteristics from unimodal biometric systems and rising and iris as a template. They said that multimodal recognition
demand. Almost all multimodal biometric systems share is more secure than single modal recognition system due to
higher matching thing are present in multimodal system.

IJISRT24MAY1763 www.ijisrt.com 2358

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1763

They used FUSION technique to processing the images. Two face and some other credentials as Name, Address, mobile
types of FUSION technique they used first one is Early fusion number, etc. We have acquired these images via simple
and Second one is Late Fusion camera (CMOS camera) and apply steganography technique.
CMOS camera is efficient to acquire the images and takes
Joseph Teena et al. [3] proposed ‘’A multimodal very lesspower.
biometric authentication scheme basedon feature fusion for
improving security in cloud environment’’ Once they have Steganography is a method or science of hiding secret
extracted the feature then they generated unique secret key by information in cover image. The interface in which the secret
fusing the traits in two stages name as FAR (FalseAcceptance information kept called Cover image.These cover image
Rate) and FRR (False Rejection Rate). They have used these travel from sender to receiver. After applying the
two features for measuring the robustness of the system. They steganography, the images converted into Stego image. This
have taken finger print, iris and palm print parallel from the stego image directly stored in the database for further
database and extracted the feature. identification. Apart from this, we can store some other
information related to person such asName, mobile number,
Walia et al. [11] proposed ‘’Secure multimodal address, department number, etc.
biometric system based on diffused graphsand optimal score
fusion’’. In this paper the author explained that most of the Below these credentials we will add during registration
multimodal biometric system has an issue of lack of
robustness and security. To overcome this problem,they have  User Data
proposed multimodal biometric system based on face, iris and
ear optimality scorelevel fusion. In this proposed model they  Name
combined unimodal feature with related key featureand to give  Address
feature transformation. In this model they have combined  Department
multiple modalities to achieved a robust and reliable  ID Card Number
multimodal biometric system.
 Face Image
III. METHOD USED  Palm Image
Firstly, in the biometric system we will verify the user
with the help of multiple biometric traits such as palm print,

Fig 1: Flow Chart Showing the Process of Steganography

IJISRT24MAY1763 www.ijisrt.com 2359

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1763

IV. RECOMMENDATIONS FOR EFFECTIVE database. Firstly, image acquis ion will be done with the help
IMPLEMENTATION of simple camera. Later analysis of acquired image will be
done. After analysis of captured image, image converted into
A. Face Image: binary (data form) formto store in the database. In the face,
Face recognition system plays very important role in having many robust information of as compare to any other
biometric system. Basically facial recognition system biometric recognition system.
categories into some steps to acquire and store in the

Fig 2: Process of Face Recognition

B. Palm Image: Line, Secondary Line and Wrinkles. These line gives rich
Palm print biometric system is basically new information of a person which can easily identify. Palmprint
recognition system as compare to some other biometric having more feature (ridges, delta point, minute feature,
system like as fingerprint, face, ear recognition system. principal line) than the finger print, hence palm print
Palmprint has unique recognition system and having high biometric system is advisable for bettersecurity purpose.
usability. Palmprint having three basic line which is Principal

Fig 3: Process of Palm Recogintion

C. Stego Image: no onecan suspect that it having some useful information,

Stego image is an image in which secret information are because the stego image is almost similarto the cover image.
kept. After the processing of steganography stego image The similarity of the stego image is based on the LSB concept.
generated. While processing of psalm image and face image, In the LSBconcept the single or double digit changes and in
a single image generated name as stego image. Stego image is that single or double digit the data (secret information) can be
a type of encoded information in which it can be placed in stored. Below the image showing before (cover image) and
anywhere. Stego image contain secret information in which after (stego image) image.

IJISRT24MAY1763 www.ijisrt.com 2360

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1763

Fig 4: Process of Stegonation

The above image showing that the efficiency of stego [4]. Karthika, P., R. Ganesh Babu, and K. Jayaram.
image. In the stego image there are not much changes as "Biometric based on steganography imagesecurity in
compare to cover image. This is why, because in the stego wireless sensor networks." Procedia Computer
image there are only 2 bit changing of LSB bit. Hence secret Science 167 (2020): 1291-1299.
message can be kept with the help of this technique. [5]. Mishra, Ashish. "Multimodal biometrics it is: need for
future systems." International journal of computer
V. CONCLUSION applications 3.4 (2010): 28-33.
[6]. Mustafa, Ahmed Shamil, Aymen Jalil Abdulelah, and
We will get the secret information in De-stego technique Abdullah Khalid Ahmed. "Multimodal Biometric
which we had kept secretly during verification process. In the System Iris and Fingerprint Recognition Based on
verification process of biometric, face image combined with Fusion Technique." International Journal of
palm image called stego image. This stego image stored in Advanced Science and Technology 29 (2020): 7423-
the database for further identification work. In the 7432.
identification process firstly we need to acquire the face [7]. Singh, Arun Kumar, Juhi Singh, and Harsh Vikram
image and that face image has already stored in the database Singh. "Steganography in images usinglsb technique."
along with palm image. International Journal of Latest Trends in Engineering
and Technology (IJLTET) 5.1 (2015): 426-430.
After acquiring the face image preprocessing of the [8]. Satya Bhushan verma and Saravanan Chandran.
image will be done. Identification module will work on the "Comparative Study of FAST MSER and Harris for
next stage from the preprocessing. If face image is identified, Palmprint Verification System." International Journal
then matched image (stego image) will be fetched out from of Scientific & Engineering Research 7.12 (2016):
the database to do remaining work. After fetching the face 855-858.
image (stego image) De-stego will be done. In this stage if [9]. Satya Bhushan verma, and Chandran Saravanan.
De-stego technique applied then palm image separated from "Performance analysis of various fusion methods in
the face image. Once the palm image separated from the face multimodal biometric." 2018 International
image with the help of De-stego then acquired palm image Conference on Computational and Characterization
will be matched to the separated image. Techniques in Engineering & Sciences (CCTES).
IEEE, 2018.
REFERENCES [10]. Satya Bhushan verma, and Saravanan Chandran.
"Touchless Region based Palmprint Verification
[1]. Ammour, Basma, et al. "Face–iris multimodal System." International Journal of Computer Science
biometric identification system." Electronics 9.1 and Information Security(IJCSIS) 15.4 (2017).
(2020): 85. [11]. Walia, Gurjit Singh, et al. "Secure multimodal
[2]. Chandran, Saravanan, and Satya Bhushan Verma. biometric system based on diffused graphsand optimal
"Touchless Palmprint Verification usingShock Filter, score fusion." IET Biometrics 8.4 (2019): 231-242.
SIFT, I-RANSAC, and LPD." IOSR Journal of [12]. Xiong, Qi, et al. "A modified chaotic binary particle
Computer Engineering 17.3 (2015): 01-08. swarm optimization scheme and its application in
[3]. Joseph, Teena, et al. "A multimodal biometric face-iris multimodal biometric identification."
authentication scheme based on feature fusion for Electronics 10.2 (2021): 217
improving security in cloud environment." Journal of
Ambient Intelligence and Humanized Computing
(2020): 1-9.

IJISRT24MAY1763 www.ijisrt.com 2361