BIOMETRIC STEGANOGRAPHY: A NEW APPROACH USING HAND … · 2017-03-04 · Data hiding is one of best topic in secret communication. A lossless data hiding technique using LSB in images

@IJRTER-2016, All Rights Reserved 96

BIOMETRIC STEGANOGRAPHY: A NEW APPROACH USING

HAND GEOMETRY

Priya Yankanchi1, Shanmukhappa A. Angadi

2

1Department of Computer Science and Engineering,

Visvesvaraya Technological University, Centre for PG Studies, Belagavi, Karnataka, India 2Department of Computer Science and Engineering,

Visvesvaraya Technological University, Centre for PG Studies, Belagavi, Karnataka, India,

Abstract— With the growth of data communication over computer network, the security of

information has become a major issue and thus data hiding technique has attracted many people

around the globe. This paper presents a new framework for secure data hiding by combining

steganography technique with biometrics traits. Steganography is the art and science of hiding the

secret data in to the cover file in such a way that no one apart from sender and intended recipient,

suspects the secret message. Here image steganography technique is used, in which secret data is

embedded in the image file which acts like cover file. Traditional steganographic techniques uses

sequential LSB data embedding method, in which secret data is going to embed in the least

significant bit position of cover image pixels in sequential manner, this method is easily vulnerable

to the attacks and attacker can easily get know to the data embedded position because data is

embedded in sequential pixel position in least significant bits. Hence to overcome from this problem

this paper proposes new framework called “Biometric steganography”. In this proposed framework

steganography is implemented by utilizing one of the biometric traits i.e. Hand geometry. Here

features are extracted from hand images of individuals and by utilizing these features one unique key

is generated. This generated key is employed to find a particular pixel position from which, to embed

the secret data in least significant bits of the cover image. Stego image is generated which contains

secret data, and embedded secret message is extracted from the stego image by utilizing generated

unique key, during information extraction. This approach gives a double layered data security.

Keywords— Biometric Steganography, Data Hiding, Key Generation, Hand Geometry, Secret Data.

I. INTRODUCTION Nowadays the security has become one of the essential issues in the case of secure data transmission.

We should take care that the transmission of a data won’t get altered by any kind of network attacks

and secure transmission of data won’t susceptible to any kind of malicious attacks while the data is

in transmission mode.

Steganography techniques are used to conceal the secret data and to protect the confidential data.

Steganography is the art of invisible communication by concealing information inside other

information. The term steganography is derived from Greek word which means “covered writing”.

Steganography plays an important role in information security. The goal of Steganography is to

avoid drawing suspicion to the existence of a hidden message. This approach of information hiding

technique has recently become popular in a number of application areas.

This paper presents new framework called “Biometric steganography” which gives double layered

security for data which has been embedded in the cover file. “Biometric Steganography” is an art of

hiding the data in to another data by using some of the biometric traits”. Biometrics is metrics related

to the human characteristics and it automatically recognizes individuals based on “physical or

behavioral” characteristics.

International Journal of Recent Trends in Engineering & Research (IJRTER)

Volume 02, Issue 09; September - 2016 [ISSN: 2455-1457]


In this frame work steganography technique is used in combination with biometric traits.

Steganography is the technique of concealing the secret data in to the one of the cover file, so that no

one can suspect the data apart from the sender and receiver.

Here image steganography strategy has been used in which the secret data is going to embed

in the cover image. In this proposed system new embedding technique is implemented in which

secret data is secretly embedded in cover image based on key, unique key is generated by using

extracted hand features and this key identifies particular pixel of cover image in which data is going

to embed. First hand image has been taken out from CASIA database and taken image is pre-

processed by cropping unwanted part of the hand image after this process by utilizing the boundary

tracing algorithm hand features being extracted and valley and tip point are located using Euclidean

distance, by using these extracted features a unique key has generated, this key identifies a particular

pixel of cover image in which secret data embeds in the least significant bits position of that selected

pixel. Stego image is generated which contains secret data within it, from this stego image data being

extracted by performing reverse process of embedding, secret data is extracted using the same hand

key which has been used while embedding. Here we have taken 25 individual’s hand images from

CASIA data base for testing and here system performance is analyzed by checking how the proposed

system works for different message lengths and the distortion of original cover image and stego

image are calculated using PSNR and MSE. As the message length increases the system’s response

time increases, MSE value increases and PSNR value decreases.

This paper is organized in to 5 sections. In section II literature survey has done which are related to

steganography techniques and hand geometry. Section III presents the proposed methodology.

Experiment results are discussed in section IV. Section V presents the conclusion.

II. LITERATURE SURVEY This section gives review of the work done by researchers on steganography technique and hand

geometry.

Arun Kumar Singh et.al. [1]. In this paper, a novel data-hiding technique based on the LSB technique

of digital images is presented. Data hiding is one of best topic in secret communication. A lossless

data hiding technique using LSB in images is presented in this paper. LSB data hiding technique

does not affect the visible properties of the image. This paper deals with hiding text in an image file

using Least Significant Bit (LSB) technique. The LSB algorithm is implemented in spatial domain in

which the payload bits are embedded into the least significant bits of cover image to derive the stego-

image.

Souvik Bhattacharyya et.al. [2]. The biometric frame work is at risk to range of assaults. These

assaults are expected to bypass the security framework or to suspend the typical working. This paper

proposes a new safety concept that has been built up by making the framework more secure with the

support of Steganography along with “Biometric safety”. Here the biometric data has been inserted

in to a skin tone part of an image with the assistance of proposed steganographic method.

Rahul Joshi et.al. [3]. This paper introduces the concept of steganography using “LSB METHOD”.

Least significant bit (LSB) insertion is a common and simple approach to embed information in an

image file. In this method the LSB of a byte is replaced with data bits.

Nader A. Rahman Moham [4]. This paper proposes a novel encryption technique with password

security in view of an upgraded form of multimodal hand geometry validation which utilizes 3

authentications layers based on “geometry verification”, “digital Steganography”, and “password




authentication”. This paper explains about the biometric authentication system by using

Steganography technique.

Poonam Rathi et.al. [5]. This paper proposes a system for “Hand geometry” Feature extraction

utilizing hand outline matching. Geometric estimations of the human hand have been utilized for

“identity-authentication” as part of various business frameworks.

Nidhi Saxena et.al. [6]. This paper reviews a study about personal authentication utilizing hand

geometry. Hand geometry is used as a part of this exploration comprises of the lengths and widths of

fingers and width of a palm.

Further papers [7-10] discuss various implementations of the biometric steganography and general

stego techniques. The ensuing section describes in detail the methodology proposed in this work.

III. PROPOSED METHODOLOGY In this paper new framework is proposed i.e. “Biometric steganography using hand geometry

biometrics”. Here steganography technique is implemented along with using one of the biometric

trait i.e. hand geometry. Individual’s hand images are captured from those hand geometric features

are extracted, these extracted features are used to generate unique key for deciding the secret data

embedding position in cover image. Secret data is going to embed in the particular pixel position

which is decided by the generated key; here key selects one particular pixel in that pixel’s least

significant bit the data is going to embed. And extraction of data is done in reverse process of

embedding and here the key which has used in embedding also used while extracting. This leads to

the highest security for the embedded data.

Figure 1. Block diagram of biometric steganography.

The proposed system has implemented in following important phases.

� Feature Extraction

� Key generation

� Data Embedding

� Data Extraction

A. Feature Extraction: Here hand geometric features are extracted by following 3 steps.

• Image Acquisition:

Here in this proposed system CASIA database is used for hand images. We have used 25 hand

images for feature extraction. CASIA database involves 5,502 hand images which are captured from

312 subjects. Proposed system uses CASIA database for hand images which is already available

hence image acquisition stage is not necessary in actual. But normally in this image acquisition phase




system consist of “CCD digital camera”, “light source”, “black flat surface”. User normally places

his hand on black flat surface and image is captured by the CCD digital camera.

• Image preprocessing:

We have acquired the hand image which is colored image; we have to convert it in to the black and

white image and unnecessary part of the captured hand image has cropped. “Median filter” had used

to remove the background noise of the image. There is specific intensity between hand and

background which we can observe because of the black colored background. Thus histogram of the

hand image is “bimodal”. The hand image can easily changed over to the “binary image” by

thresholding. The threshold value consequently figured by utilizing “Otsu” method. Hand image

outline can be smoothened by applying some of the morphological operations.

• Feature Extraction:

After preprocessing stage the feature extraction of hand has performed by using boundary tracing

algorithm and some of the calculations are used to find valley and tip points. Euclidean distance is

calculated between tip and valley points.

B. Key Generation:

• Key has generated in four ways

- From extracted features tip and valley points are located

- Pair wise distance has calculated using Euclidean distance.

- Variance has calculated for pair wise distance

- And lastly unique key has generated using variance result.

Pair wise distance calculated using Euclidean distance

Given an m-by-n data matrix X, which is treated as m (1-by-n) row vectors x1, x2... xm, the various

distances between the vector xr and xs are defined as follows:

Variance The mathematical formula to calculate the variance is given by:

σ

2 = variance

∑ (X - µ)2 = The sum of (X - µ)

2 for all data points

Standard Deviation Formula The standard deviation formula is similar to the variance formula. It is given by:

σ = standard deviation

xi = each value of dataset

x (with a bar over it) = the arithmetic mean of the data (This symbol will be indicated as mean from

now)

N = the total number of data points

∑ (xi - mean)^2 = The sum of (xi - mean)^2 for all data points.

Unique 4 digit key generation

Key= (Most Significant 2 Digits of p1) * 100 + (Least Significant 2 Digits p2)




Where

P1 = variance of Hand points pairs wise distance

P2 = standard deviation of Hand points pairs wise distance

• From this unique key we will get unique pixel position to embed the secret data instead of

embedding the data in sequential pixels of cover image.

• This technique gives more security and it’s not possible for opponents to find the accurate pixel

position where the secret message has been embedded.

C. Data Embedding: In this proposed system new data embedding technique has implemented. Here key is generated

using hand geometric features and key determines one particular pixel position from that pixel the

secret data starts embedding in least significant bits positions. This leads to the highest security for

the embedded data.

Steps for LSB data insertion-

• First carrier image or cover image has been read and converted in to array of bits.

• Then the secret message which is in the form of byte/characters are converted in to the “ASCII”

values and then ASCII values are converted in to array of bits.

• A unique key which has generated from hand geometric features based on this we starts

embedding the secret data

• Here key will gives unique pixel position of cover image to embed the data least significant bit of

particular image pixel of cover object. Here one pixel is equals to one byte.

• Stego-image has been generated which contains secret message embedded within cover image.

D. Data Extraction:

• The extracting of the embedded data is done in opposite direction of hiding process.

• Here embedded secret data has extracted from stego-image

• While extracting secret data the hand key is used. while embedding the data we have used key to

select the embedding position in cover image so while extracting also we need that key to extract

secret data

• This gives complete security to the embedded data in the cover image.

IV. RESULTS AND DICUSSIONS This section describes about the testing and analysis of “Biometrics steganography”. Here in this

proposed system 25 hand images are tested and analyzed which are taken from CASIA database


Figure 2 . Secret message entering.

Figure 2 depicts that in this proposed system user has to enter the s

embed in the cover image.

Figure 3.

In figure 3 hand of the person 1 has used to gene

generated from remaining 24 individuals further for embedding


Volume 02, Issue 09; September - 2016

Figure 2 . Secret message entering.

this proposed system user has to enter the secret message which is going to

Selection of hand for feature extraction.

1 has used to generate the key for embedding. And

generated from remaining 24 individuals further for embedding the secret data in to the cover image.


2016 [ISSN: 2455-1457]

ecret message which is going to

rate the key for embedding. And also the key is

the secret data in to the cover image.


Figure 4. Generation of key from ex

In figure 4 tip and valley points are calculated and hand key is generated from extracted feature.

Figure 5. Stego image generation.

Figure 5 depicts that after selecting cover image, the secret data has been embedded in to the

image and stego image has generated i.e. the secret data embedded within cover image



Figure 4. Generation of key from extracted hand features.


Figure 5. Stego image generation.

fter selecting cover image, the secret data has been embedded in to the

image and stego image has generated i.e. the secret data embedded within cover image


2016 [ISSN: 2455-1457]


fter selecting cover image, the secret data has been embedded in to the cover

image and stego image has generated i.e. the secret data embedded within cover image.


Figure 6. Extracted secret data.

Figure 6 depicts that the secret data got extracted from stego image after selecting the hand image

which has generated unique key which has used while embedding process.

The proposed framework’s performance is analyzed by examining the distortion/similarity

statistically. MSE and PSNR are used to calculated distortion between original and stego image.

Table 1. Performance

Message Length Response time (Sec)

10

100

190

280

370

460

550

640

730

820

910

1000

Table 1 shows that the performance analysis of proposed system. Here the distortion of original

cover image and stego image are calculated using PSNR and MSE. As the message length increases

the system’s response time increases, MSE value increases and PSN

V.Nowadays data security is one of the critical issues while transferring the secret information from

one to another place by utilizing the internet, because unauthorized people can hack the secret

information in between data transfer and make it futile or acquires the secret information which is

not intended to him.



Figure 6. Extracted secret data.

the secret data got extracted from stego image after selecting the hand image

ey which has used while embedding process.



Performance analysis of proposed system.

Response time (Sec)

PSNR

1.073 89.31

1.616 76.13

2.167 73.11

2.790 71.51

3.352 70.23

3.992 69.30

4.663 68.51

5.224 67.96

5.772 67.44

6.330 66.80

6.892 66.34

7.458 65.97



the system’s response time increases, MSE value increases and PSNR value decreases.

V. CONCLUSION Nowadays data security is one of the critical issues while transferring the secret information from


ransfer and make it futile or acquires the secret information which is


2016 [ISSN: 2455-1457]

the secret data got extracted from stego image after selecting the hand image



MSE

0.000076

0.001587

0.003174

0.004593

0.006165

0.007645

0.009171

0.010406

0.011734

0.013596

0.015106

0.016434



R value decreases.

Nowadays data security is one of the critical issues while transferring the secret information from


ransfer and make it futile or acquires the secret information which is




Hence to overcome from this problem the proposed system provides double layered data security by

utilizing steganographic technique with biometrics. “Biometric steganography”, is one of the new

frame work which has been implemented in this paper. Biometric steganography is the technique in

which data is being hidden in the cover image by utilizing hand geometric extracted key. In this

proposed system image steganography is used in which secret text data is embedded in the cover

image using a key which has generated from hand geometric extracted features. In traditional LSB

technique data has been embed in least significant bit positions, normally in this technique the

attacker can visualize the embedded secret data positions so to overcome from this problem the

proposed system implements the new data embedding technique in which data is going to embed by

utilizing a unique key which has been generated by extracted hand geometric features.

The key selects particular pixel position starting from this pixel, data is being embedded. In this

proposed system 25 individual’s hand image is taken from CASIA database for testing, from these

hand images features are extracted and unique key is generated respectively. Key is used to

determine secret data embedding positions. Here system performance is analyzed by checking how

the proposed system works for different message lengths and the distortion of original cover image

and stego image are calculated using PSNR and MSE. As the message length increases the system’s

response time increases, MSE value increases and PSNR value decreases.

This proposed system can be further improved by embedding huge amount of data in cover image

without much loss in quality of cover image. And other biometric traits can be used to generate key

in simple way which gives robustness to the system.

REFERENCES 1. Arun Kumar Singh, Juhi Singh and Dr. Harsh Vikram Singh “Steganography in Images using LSB Technique”,

International Journal of Latest Trends in Engineering and Technology (IJLTET) Volume 5, Issue 1, January 2015.

2. Souvik Bhattacharyya, Indradip Banerjee, Anumoy Chakraborty and Gautam Sanyal “Biometric Steganography

Using Variable Length Embedding”, International Journal of Computer, Electrical, Automation, Control and

Information Engineering, Volume 8, 2014.

3. Rahul Joshi, Lokesh Gagnani and Salony Pandey “Image Steganography with LSB”, International Journal of

Advanced Research in Computer Engineering & Technology (IJARCET) Volume 2, Issue 1, January 2013.

4. Nader A. Rahman Mohamed “Multi-Modal(Hybrid) More Efficient and highly Secured Enhanced Geometry

Authentication Using 3 Steps Authentication by Means of Biometrics, Steganography and Encryption”, Journal of

Image Processing & Pattern Recognition Progress(JoIPPRP) Volume 1,Issue 2, 2014.

5. Poonam Rathi and Dr. Sipi Dubey “Hand Geometry Recognition System Using Feature Extraction”, International

Journal of Advanced Research in Computer Engineering and Technology (IJARCET) Volume 2, Issue 6, June

2013.

6. Nidhi Saxena, Vipul Saxena, Neelesh Dubey and Pragya Mishra “HAND GEOMETRY: A New Method for

Biometric Recognition”, International Journal of Soft Computing and Engineering (IJSCE) Volume 2, Issue 6,

January 2013.

7. Abbas Cheddad, Joan Condell, Kevin Curran and Paul Mc Kevitt “Biometric Inspired Digital Image

Steganography”, 15th Annual IEEE International Conference and Workshop on the Engineering of Computer

Based Systems, 2008.

8. Amritha.G and Meethu Varkey “Biometric Steganographic Technique Using DWT and Encryption”, International

Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE) Volume 3, Issue 3,

March 2013.

9. N.Lavanya, V.Manjula and N.V. Krishna Rao “Robust and Secure Data Hiding in Image Using Biometric

Technique”, International Journal of Computer Science and Information Technologies (IJCSIT) Volume 3 , 2012.

10. Saurabh Singh and Gaurav Agarwal “Use of image to secure text message with the help of LSB replacement”,

International Journal of Applied Engineering Research, Dindigul, Volume 1, No1, 2010.

BIOMETRIC STEGANOGRAPHY: A NEW APPROACH USING HAND … · 2017-03-04 · Data hiding is one of best topic in secret communication. A lossless data hiding technique using LSB in images

Documents