See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/331544868

A Robust and Secured Image Steganography using LSB and Random Bit
Substitution

Article · February 2019

CITATIONS READS

18 913

3 authors:

U A Md Ehsan Ali Md Sohrawordi

Hajee Mohammad Danesh Science and Technology University Hajee Mohammad Danesh Science and Technology University
12 PUBLICATIONS 26 CITATIONS 11 PUBLICATIONS 27 CITATIONS

SEE PROFILE SEE PROFILE

Md. Palash Uddin

Deakin University
54 PUBLICATIONS 473 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Data Security View project

Remote Sensing Satellite Image Processing View project

All content following this page was uploaded by U A Md Ehsan Ali on 06 March 2019.

The user has requested enhancement of the downloaded file.

American Journal of Engineering Research (AJER) 2019
American Journal of Engineering Research (AJER)
E-ISSN: 2320-0847 p-ISSN: 2320-0936
Volume-8, Issue-2, pp-39-44
www.ajer.org
Research Paper Open Access

A Robust and Secured Image Steganography using LSB and

Random Bit Substitution
U. A. Md. Ehasn Ali1, Md. Sohrawordi1, Md. Palash Uddin1
1
(Department of Computer Science and Engineering, Hajee Mohammad Danesh Science and Technology
University, Dinajpur-5200, Bangladesh)
Corresponding Author: U. A. Md. Ehasn Ali

ABSTRACT :In the era of information, secret communication is one of the most important issues in today's
communication systems. The development of digital communication with data secrecy becomes one of the main
purposes of the researchers. Among the data security techniques, steganography is an approach to hide the
existence of secret message in a carrier without being apprehended by the intruders. At present, it is important
to hide data as well as develop mechanisms by transporting them in such a way that the intruders cannot
understand the message. There are various data hiding methods that use different digital media. Digital images
are the most popular among the carrier formats because of their frequency on the Internet. In this paper, a new
approach of image steganography with least significant bit substitution is proposed where the information
embedded in the random bit position of a pixel. The experimental result signifies the importance of the proposed
method.
KEYWORDS: steganography, LSB substitution, data hiding, embedded message, retrieved message.
----------------------------------------------------------------------------------------------------------------------------- ----------
Date of Submission: 22-01-2019 Date of acceptance: 07-02-2019
----------------------------------------------------------------------------------------------------------------------------- ----------

I. INTRODUCTION
Communication in secret ways with others has always been one of the well-known glitches from the
ancient times. Nowadays, although cryptography and steganography are used for secured communication but
they are different in nature [1]-[6]. As cryptography reforms a message, it cannot be understood whereas
steganography hides the message so that it cannot be seen. The aim of steganography is to embed the
communication content in a public cover media [2], [3], [6]. As a result, the presence of concealed message can
be hidden. Digital steganography is the science of transferring information with secret message embedded in it.
Thus, steganography is a form of security technique where the existence of a message is kept hidden between
the sender and the intended recipient. Steganography is classified into 3 categories [2], [8]-[10],
• Pure steganography with no stego key. Here, the communicating parties assume that no other party is aware
of the communication.
• Secret key steganography with the stego key. Here, the communicating parties exchange the key prior to
communication. This is most susceptible to interruption.
• Public key steganography with a public key and a private key. Here, the both keys are used for secure
communication.
In general, Steganography focuses on hiding the secret messages in carriers with unnoticed and less
attractive way. Almost all digital file formats can be used as a carrier, but the suitable formats are those which
have a high degree of redundancy. There are various methods and algorithms of hiding data in different types of
digital file formats. There are many suitable steganographic techniques which are being used to obtain security
depending on the type of the carrier [2], [8]-[10].
• Image Steganography: When an image used as a cover object in steganography, it is known as image
steganography. Generally, pixel intensities of the image are used to hide the information in these
techniques.
• Network Steganography: Taking network protocol, such as TCP, UDP, ICMP, IP etc. as cover object, is
known as network protocol steganography. Here, steganography can be achieved by using unused header
bits of network protocols.

www.ajer.org Page 39
American Journal of Engineering Research (AJER) 2019
• Video Steganography: In video Steganography, the carrier for hidden information is video (combination of
pictures). To hide the information in each of the images in the video, normally, discrete cosine transforms
(DCT) technique is used. Various video formats such as H.264, Mp4, MPEG, AVI etc. are used in video
steganography.
• Audio Steganography: While using audio as a carrier for hiding information, it is called audio
steganography. Because of popularity of voice over IP (VOIP), it has become very important medium.
Digital audio formats such as WAVE, MIDI, AVI MPEG or etc. are used for this steganography.
• Text Steganography: To achieve information hiding in text steganography, different techniques such as
capital letters, white spaces, number of tabs, just like Morse code and etc. is used.

II. IMAGE STEGANOGRAPHY

In today’s world, almost in every page on the internet, digital images can be found due to the easiest
distribution. Therefore, image files are the most commonly used environments for steganography applications,
although they vary according to the formats used [2], [8]-[10]. The human visual system has a weakness that is;
it has a low sensitivity towards random pattern changes and luminance. The human eye cannot detect the small
changes in color or patterns and due to this weakness, text or graphic files can be inserted into the carrier image
without being detected. For example, an image file with an encoded text stored in it results another image file,
called “stego image”, that is physically and visually not the same from the original. Hence, an eavesdropper
only sees a picture that is transferred between the two parties when a communication take place between them,
but they are unaware of the hidden messaging actually takes place by this picture. However, an image is a
collection of individual points, referred to as pixels that constitute different light intensities in different areas of
the image. In an image, the pixels are presented horizontally row by row. Different color schemes use different
number of bits for each pixel, called the bit depth. The smallest bit depth in current color schemes is 8, that is, to
represent the color of each pixel 8 bits are used (corresponding to 2^8 = 256 colors). 8 bits are used in
monochrome and grayscale images for each pixel and are able to display 256 different shades of gray. Digital
color images use 24-bit for each pixel and use the RGB color model, also known as true color. Different
methods can be used to hide information in images. These methods can be categorized under two headings,
considering the data they use during embedding.
1. Spatial / Image Domain Technique
2. Frequency / Transform Domain Technique
In Spatial Domain or Image Domain, the pixels of the image file are directly changed for embedding
secret data in it. An example of this technique is the Least Significant Bit Insertion (LSB) method, which is
commonly used. In Frequency Domain or Transform Domain, the cover image is transformed in frequency
domain from spatial domain and then its frequencies are used to hide the secret data. After hiding, the object is
again transformed into spatial domain. An example of the Transform Domain technique is the discrete wavelet
transform (DWT) implementation that uses a discrete set of the wavelet scales and translations. To evaluate the
quality of the method used for information hiding, this research takes on a number of Standard parameters for
the purpose of measuring the quality of images. The most common measures used to determine the quality of
image are Peak Signal to Noise Ratio (PSNR) and Mean Squared Error (MSE) metrics. MSE is to estimate the
mean of the squares of the error between stego image and the original image [11].
1 𝑀 𝑁
𝑀𝑆𝐸 = ∑ ∑ ( 𝑋𝑖𝑗 − 𝑋 ′ 𝑖𝑗 )
𝑀𝑁 𝑖=1 𝑗=1
PSNR is often used a s a quality measurement to determine the degradation in the embedding image with
respect to the cover image, that is, the difference between the original and the stego image [11].
𝐼2
𝑃𝑆𝑁𝑅 = 10𝑙𝑜𝑔10
𝑀𝑆𝐸
Where 𝑋𝑖𝑗 is the pixel in the original image (cover image) in ith row and jth column, 𝑋’𝑖𝑗 is the pixel in
the stego image in ith row and jth column, 𝑀𝑁 is the size of the image where M is the height and N is the width
and 𝐼 is the range pixel value. For 8-bit images, 𝐼=255.

III. RELATED WORK

Due to its plainness and non-detectability, Least Significant Bit (LSB) steganography has been
commonly used to hide information within image. Different authors used the simple LSB techniques to hide
data by replacing the pixels with secret data bits. Zhou, Gong, Fu and Jin proposed a method that was built up
by combining Rivest-Shamir-Adleman (RSA) cryptology and LSB steganography. Compared to the standard
LSB substitution technique, this method achieved a 5.8% increase in PSNR value [12]. Akhtar, Khan and Johri
introduced a method with two different LSB methods were suggested based on bit inversion. In this paper, 3
different images were used to hide 6 different messages. As a result of the experiments carried out, the PSNR

www.ajer.org Page 40

www.ajer.org
American Journal of Engineering Research (AJER) 2019
value was improved by 5.92% and 15.8% in first and second proposed method respectively [13]. A. Singh and
H. Singh proposed a new LSB steganography method where 2-2-4 message bits are embedded in the R-G-B
channels, respectively. In the study, applying the method on 2 different images, PSNR value was improved by
36.44% and 64.54%, respectively [14]. To hide information in grayscale and color images, Goyal, Ramaiya and
Dubey proposed 1-2-4 LSB method with the RSA algorithm to encrypt the message. 4 different images were
used to test the method and up to 41.48% improved PSNR value was achieved [15]. Sugathan proposed a new
LSB steganography algorithm where different embedding direction of message bits was used. Implementing the
method on 10 different images, a 1.32% improvement in PSNR value has been achieved compared to the
classical LSB method [16]. On the other hand, Kaur and Kochhar introduced two different LSB and DCT
techniques were used to perform steganography. Comparing the PSNR values with previous works the method
showed a good result and the security was improved by using DCT [17].

IV. THE PROPOSED METHOD

Sometimes an image with an embedded file inside can be wide-open to many threats that pose a risk to
the files to be hidden. A new method is proposed to preserve the secrecy of this information and prevent it from
being exposed to any threat that may lead to its revelation. The 24-bit color images with RGB color model use 8
bits each for red, green and blue value to represent a pixel. The proposed method uses a modified LSB technique
where the random position of the RGB binary representation of the color image is used to hide message bits.
The overall method is shown in Figure 1.

Figure 1: Proposed method

The embedding process of the proposed method is as follows:

1. Convert the message into binary form.
2. Generate random bit position from bit 1 to 7 for each R, G and B values of each pixel of the image.
3. Perform XOR operation of randomly chosen position bit with message bit.
4. Substitute the result with the LSB of the R, G or B values.
In this method, the secret message is converted into binary form to embed it in the image. A Pseudo
Random Number Generator (PRNG) is used to generate random bit position from 7 most significant bits for
each R, G and B values of each pixel of the image. The 1 st message bit and the randomly chosen position bit of
R value are used for Exclusive-OR operation and the result is substituted with the least significant bit. The
process follows with the 2nd message bit and the randomly chosen position bit of G value in next and the 3 rd
message bit and the randomly chosen position bit of B value in next and so on. For example, if the message is

www.ajer.org Page 41

www.ajer.org
American Journal of Engineering Research (AJER) 2019
01000101 and the randomly chosen position is 5, 2, 6, 4, 5, 1, 7 and 4 then the encoding process is illustrated in
Table 1.

Table 1: Embedding process

Pixel Binary with randomly chosen Message XOR of randomly chosen Substitute the result with
values position from 1 to 7 bit bits position bit with message bit the LSB position
R 11001001 0 1⨁0=1 11001001
G 01111110 1 1⨁1=0 01111110
B 00100100 0 1⨁0=1 00100101
R 11110111 0 1⨁0=1 11110111
G 10010100 0 0⨁0=0 10010100
B 00011110 1 0⨁1=1 00011111
R 11110111 0 1⨁0=1 11110111
G 10010100 1 1⨁1=0 10010100

The extraction process of the proposed method is as follows:

1. Generate random bit position from bit 1 to 7 for each R, G and B values of each pixel of the image.
2. Perform XOR operation of randomly chosen position bit with LSB.
3. Collect the result as that is the message bit.
4. Retrieve the message from the result by converting.
To retrieve the message from the image, first we need to use the same Pseudo Random Number
Generator to generate random bit position. Then Exclusive-OR operation is performed in randomly chosen
position bit of R value and least significant bit of R value to extract the 1 st message bit and the process goes on.
Finally, with all the message bits, the secret message is retrieved by converting them. The process is illustrated
in Table 2 to extract the message of the example considered.

Table 2: Message retrieving process

Pixel values Binary with randomly chosen position from XOR of randomly chosen position bit Message
1 to 7 bit with LSB bits
R 11001001 1⨁1=0 0
G 01111110 1⨁0=1 1
B 00100100 1⨁1=0 0
R 11110111 1⨁1=0 0
G 10010100 0⨁0=0 0
B 00011110 0⨁1=1 1
R 11110111 1⨁1=0 0
G 10010100 1⨁0=1 1

V. EXPERIMENTAL RESULT
The proposed steganography method has been implemented in MATLAB (R2014a) using the mostly
available three 24-bit images (Lena, pepper and baboon) with size 512×512as shown in Figure 2and the MSE,
SNR and PSNR have been measured. The message to embed in the entire experiment is "HelloWorld".

Lena Peppers Baboon

Figure 2: Cover images used in the experiment

The results of the implemented method are illustrated in Table 3:

Table 3: The measures of MSE, SNR and PSNR
Cover Image MSE SNR PSNR
Lena 0.00004196 85.0479 90.1855
Peppers 0.00002670 84.1735 90.0977
Baboon 0.02022934 56.1544 61.4840

www.ajer.org Page 42

www.ajer.org
American Journal of Engineering Research (AJER) 2019
The result shows that the proposed method have the same PSNR value as the standard LSB method,
but the secret message is hidden in the random bit position of the pixel value in the proposed method whereas
the message is hidden in least significant bit of the pixel value in standard LSB. Due to this random bit
substitution, the attackers cannot retrieve the message although they detect the presence of the secret message in
the image. In addition, to enhance the security we can add an encryption technique for the text. Hence, the
proposed method provides a robust image steganography for secured communication.

VI. CONCLUSION
To communicate digitally with each other, and to share and exchange electronic documents among them,
computer-based communications are main and essential in modern life. In order to secured and protected
communication over Internet, the proposed method of image steganography hides the message bit in random
position of the pixel bits and hides a reference like standard LSB technique. Thus, it hides the reference not the
real data. The experimental measures PSNR and MSE for the proposed technique show that the method provides
good performance compared with the standard LSB in regards to invisibility and robustness.

REFERENCES
[1]. Uddin,M.P., Marjan, M.A.,Sadia, N.B.,Islam, M.R.: Developing an Efficient Algorithm to Combine Cryptography and
Steganography Based on ASCII Conversions and Cyclic Mathematical Function, 3rd IEEE International Conference on Informatics,
Electronics & Vision, Dhaka, Bangladesh,May 23-24(2014).
[2]. Islam, M.R.,Siddiqa, A., Uddin,M.P., Mandal, A.K., Hossain, M.D.: An efficient filtering based approach improving LSB image
steganography using status bit along with AES cryptography, 3rd IEEE International Conference on Informatics, Electronics &
Vision, Dhaka, Bangladesh, May 23-24(2014).
[3]. Uddin,M.P.,Saha, M.,Ferdousi, S.J., Afjal, M.I., Marjan, M.A.: Developing an Efficient Solution to Information Hiding through
Text Steganography along with Cryptography, 9th IEEE International Forum on Strategic Technology, CUET, Bangladesh, October
21-23, (2014).
[4]. Uddin,M.P., Afjal, M.I., Nitu, A.M., Islam, M.M., Marjan, M.A.: Development of a Symmetric-key Cryptographic Algorithm based
on ASCII and Radix Conversions, Institutional Engineering and Technology (IET), 5(1):19-24, April, (2015).
[5]. Basak, P., Arjuman, L., Nitu, A.M., Afjal, M.I., Uddin,M.P., Rabbi, M.F., Islam, M.R.: A Modified Blind Steganalysis Method
Based on the Moments of Characteristic Function, HBRP Advancement in Software Engineering and Testing, Volume 1 Issue 2,
pp. 1-9 (2018).
[6]. Sultana, S., Khanam, A., Islam, M.R., Nitu, A.M., Uddin,M.P., Afjal, M.I., Rabbi, M.F.: A Modified Filtering Approach of LSB
Image Steganography Using Stream Builder along with AES Encryption, HBRP Recent Trends in Information Technology and its
Applications, Volume 1 Issue 2, pp. 1-10 (2018).
[7]. Morkel, T.,Eloff, J.H.P., Olivier,M.S.: An Overview of Image Steganography, Proceedings of the Fifth Annual Information Security
South Africa Conference,Sandton, South Africa (2005).
[8]. Saha,A.,Halder, S., Kollya, S.: Image steganography using 24-bit bitmap images, 14th International Conference on Computer and
Information Technology, 56-60, (2011).
[9]. Karim,S.M.M., Rahman,M.S., Hossain, M.I.: A New Approach for LSB Based Image Steganography using Secret Key,
Proceedings of 14th IEEE International Conference on Computer and Information Technology, 286 – 291 (2011).
[10]. Wayner, P.: Disappearing Cryptography: Information Hiding: Steganography & Watermarking, ELSEVIER, 3rd Edition(2008).
[11]. Sumathi, C.P., Santanam, T., Umamaheswari, G.: A Study of Various Steganographic Techniques Used for Information Hiding”,
International Journal of Computer Science & Engineering Survey (IJCSES) Vol.4, No.6 (2013).
[12]. Zhou, X., Gong, W., Fu, W., Jin, L.: An Improved Method for LSB Based Color Image steganography Combined with
Cryptography, ICIS 2016, Okayama, Japan, IEEE June 26-29 (2016).
[13]. Akhtar, N., Khan, S., Johri, P.: An Improved Inverted LSB Image Steganography, IEEE International Conference on Issues and
Challenges in Intelligent Computing Techniques (ICICT) (2014).
[14]. Singh, A., Singh, H.: An Improved LSB based Image Steganography Technique for RGB Images, IEEEInternational Conference on
Electrical, Computer and Communication Technologies (ICECCT) (2015).
[15]. Goyal, S., Ramaiya, M., Dubey, D.: Improved Detection of 1-2-4 LSB Steganography and RSA Cryptography in Color and
Grayscale Images, IEEE International Conference on Computational Intelligence and Communication Networks (2015).
[16]. Sugathan, S.: An Improved LSB Embedding Technique for Image Steganography, IEEE International Conference on Applied and
Theoretical Computing and Communication Technology (2016).
[17]. Kaur, G., Kochhar, A.: A steganography implementation based on LSB & DCT. International Journal for Science and Emerging
Technologies with Latest Trends, 4(1), pp.35-41 (2012).

U. A. Md. Ehasn Ali (ehsan_cse@hstu.ac.bd) received his B. Sc. degree in Computer Science and
Engineering from Hajee Mohammad Danesh Science and Technology University, Dinajpur,
Bangladesh in 2013. His main working interest is based on Image Processing, Expanding the
Applications of Artificial Intelligence, Machine Learning, Data Mining, Data Security
etc. Currently, he is working as a lecturer in Dept. of Computer Science and Engineering in Hajee
Mohammad Danesh Science and Technology University, Dinajpur, Bangladesh. He has several
scientific research publications in various aspects of Computer Science and Engineering.

www.ajer.org Page 43

www.ajer.org
 American Journal of Engineering Research (AJER) 2019
Md. Sohrawordi (mdsohrawordicse@gmail.com) is working as a lecturer in Dept. of Computer
Science and Engineering in Hajee Mohammad Danesh Science and Technology University,
Dinajpur, Bangladesh. He received his B. Sc. degree in Computer Science and Engineering from
Hajee Mohammad Danesh Science and Technology University, Dinajpur, Bangladesh in 2013.
Now, he is pursuing M. Sc. degree in Computer Science and Engineering from Rajshahi
University of Engineering & Technology (RUET), Rajshahi, Bangladesh. His main working
interest is based on Image Processing, Artificial Intelligence, Data Mining, Mobile
Networks,cryptographyetc.He has several scientific research publications in various aspects of
Computer Science and Engineering.

Md. Palash Uddin (palash_cse@hstu.ac.bd), a member of IEEE, is presently serving as an

Assistant Professor in department of Computer Science and Engineering in Hajee Mohammad
Danesh Science and Technology University, Dinajpur, Bangladesh. Previously, he was a lecturer
in the same university and in department of Computer Science and Engineering at Central
Women’s University, Dhaka, Bangladesh. He has completed his M. Sc. degree from department
of Computer Science & Engineering, Rajshahi University of Engineering & Technology (RUET),
Rajshahi, Bangladesh in 2018. He received his B. Sc. degree in Computer Science and
Engineering from Hajee Mohammad Danesh Science and Technology University, Dinajpur,
Bangladesh in 2011. His focal research interest is based on the remote sensing image analysis,
artificial intelligence based application development and machine learning algorithms for data
mining. He has several national and international journal and conference publications in various
fields of Computer Science and Technology. In 2017, he received the “best paper award” on the
paper tilted as “Feature Extraction for Hyperspectral Image Classification” in the prestigious
IEEE 5th Region 10 Humanitarian Technology Conference (R10HTC) hosted by Bangladesh
University of Engineering and Technology (BUET), Dhaka, Bangladesh. In 2018, he as a co-
author of the paper entitled “Weighted-Correlation based Band Reordering Heuristics for
Lossless Compression of Remote Sensing Hyperspectral Sounder Data” received “best paper
award” in IEEE International Conference on Advancement in Electrical and Electronic
Engineering-(ICAEEE 2018) organized by Dhaka University of Engineering and Technology
(DUET), Gazipur, Bangladesh.

U. A. Md. Ehasn Ali" A Robust and Secured Image Steganography using LSB and Random Bit
Substitution " American Journal of Engineering Research (AJER), vol.8, no.02, 2019, pp.39-44

www.ajer.org Page 44

www.ajer.org
View publication stats