It3116411644

Pritish Bhautmage, Prof. Amutha Jeyakumar / International Journal of Engineering Research
and Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 3, Issue 1, January -February 2013, pp.1641-1644
Advanced Video Steganography Algorithm
Pritish Bhautmage*, Prof. Amutha Jeyakumar**, Ashish Dahatonde***
*(Department of Elecrical Engineering, VJTI, Mumbai.)
** (Department of Electrical Engineering, VJTI, Mumbai.)
*** (Department of Electrical Engineering, VJTI, Mumbai.)

ABSTRACT of bits. The cover or host is the medium in which
Data embedding is the process of the message is embedded and serves to hide the
embedding information in a data source without presence of the message. This is also referred to as
changing its perceptual quality. Several the message wrapper. It is not required that the
constraints affect this process: the quantity of cover and the message have homogeneous structure.
data to be hidden, the need for invariance of In addition, the encoder usually employs a stego-key
these data under the conditions where a host which ensures that only recipients who know the
signal is subject to distortions such as lossy corresponding decoding key will be able to extract
compression and the degree to which the data the message from a stego-message.
must be immune to interception, modification or
Cover object
removal by a third party. A new technique is
proposed in this paper for data embedding and C
extraction for high resolution AVI videos. In this
method instead of changing the LSB of the cover Stego-object
file, the LSB and LSB+3 bits are changed in Secret object
alternate bytes of the cover file. The secret Z
message is encrypted by using a simple bit M F(C, M, K)
exchange method before the actual embedding
process starts. An index can also be created for
the secret information and the index is placed in Stego-Key
a frame of the video itself. With the help of this
index, we can easily extract the secret message, K
which can reduce the extraction time. The
different techniques and advantages of video Fig 1. Digital Steganographic Encoder
steganography is discussed in this paper.

Keywords – Encryption, Index Creation, LSB The original cover file may or may not be
Technique, Stego Key. required; in most applications it is desirable that the
cover file not be needed to extract the message.
I. INTRODUCTION It requires the cryptographic decoding key
In conventional cryptography, even if the to decipher the encrypted message[3]. The
information contents are protected by encryption, requirements of any data embedding system can be
the existence of encrypted communications is categorized into security, capacity and robustness.
known. In view of this, digital steganography All these factors are inversely proportional to each
provides an alternative approach in which it other creating so called data embedding dilemma.
conceals even the evidence of encrypted messaging. We focus at maximizing the first two factors of data
Generally, steganography is defined as the art and embedding i.e., security and capacity. We have
science of communicating in a covered fashion[1][2]. attempted on videos rather than on audio and image.
It utilizes the typical digital media such as text, The objective of the paper is to develop an
image, audio, video, and multimedia as a carrier algorithm for data embedding in AVI videos.
(called a host signal) for hiding private information The proposed scheme is a data embedding
in such a way that the third parties (unauthorized method that uses high resolution digital video as a
person) cannot detect or even notice the presence of cover signal. The proposed scheme provides the
the communication. In this way, steganography ability to hide a significant quality of information
allows for authentication, copyright protection, and making it different from typical data embedding
embedding of messages in the image or in mechanisms because we consider application that
transmission of the image. A typical digital require significantly larger payloads like video-in-
steganographic encoder is shown in Figure (1). The video and image-in-video[4]. The purpose of
message is the data that the sender wishes to remain embedding such information depends on the
confidential and can be text, images, audio, video, application and the needs of the owner/user of the
or any other data that can be represented by a stream digital media.

1641 | P a g e

Data embedding requirements include the following: modifications in case a lossy compression algorithm
1) Imperceptibility: The video with data and original like JPEG is being used[6].
data source should be perceptually identical. In transform based data embedding, the cover
2) Robustness: The embedded data should survive image is transformed into another domain. Then the
any processing operation the host signal goes data is embedded in the transform coefficients. This
through and preserve its fidelity. method is highly robust and complex. The major
3) Capacity: Maximize data embedding payload. transformations used are DCT and DWT. DCT is
4) Security: Security is in the key for embedding or used in JPEG compression algorithm to transform
encryption of data. successive 8_8 pixel blocks of the image, into 64
DCT coefficients each. After calculating the
The paper is organized as follows. Related coefficients, the quantizing operation is performed.
work of the proposed technique is explained in Although a modification of a single DCT will affect
section II. The methodology of the proposed all 64 image pixels, the LSB of the quantized DCT
technique is described in section III. Section IV coefficient can be used to embed information.
concludes the paper. When information is hidden in video, the
program or person embedding the information will
II. RELATED WORK usually use the DCT method. DCT works by slightly
There are mainly three basic data changing the coefficients of each of the images in
embedding techniques for images in practice, the video, only so much that it is not noticeable by
namely Least Significant Bit (LSB) Method, the human eye. Data embedding in videos is similar
Masking and filtering and Transform based[5]. to that of data embedding in images, apart from
The primitive method is embedding in information is hidden in each frame of the video.
LSB. Although there are several disadvantages to When only a small amount of information is hidden
this approach, the relative easiness to implement it in a video, generally it is not noticeable. However,
makes it a popular method. In this method we when more information is hidden, it will be more
embed information in the LSB of pixels colours. noticeable.
The changes of LSB may not be noticeable because DWT is based on sub-band coding and is
of the imperfect sensitivity of the human eyes. On found to yield a fast computation of Wavelet
an average, only half of the bits in an image will Transform. It is easy to implement and reduces the
need to be modified to embed a secret message computation time and resources required. A 2-D
using the maximal cover size. While using a 24-bit DWT transforms an image into four sub bands: LL,
image gives a relatively large amount of space to LH, HL and HH where L and H stands for Low and
hide messages, it is also possible to use an 8-bit High. The LL sub band contains the average
image as a cover source. Because of the smaller information and the other three sub-band gives the
space and different properties, 8-bit images require a finer details of the image. Even if the three sub-
more careful approach. Where 24-bit images use bands LH, HL, HH are made zero, the LL alone can
three bytes to represent a pixel, an 8-bit image uses give the average image (an image of lower quality,
only one. Changing the LSB of that byte will result with no finer details). We can embed the message
in a visible change of colour, as another colour in image in two LSB planes of LH, HL and HH sub
the available palette will be displayed. Therefore, bands. Data is embedded in LL sub-band to avoid
the cover image needs to be selected more carefully compression losses. Human Visual System (HVS)
and preferably be in gray scale, as the human eye model points out different insensitivities among
will not detect the difference between different gray different level sub bands. More insensitive to HVS
values as easy as with different colours[4]. means that more data can be embedded without
Masking and filtering techniques, usually causing notable visual artifacts.
restricted to 24 bits or gray scale images, take a
different approach to embedding a message. These Transform-based method is found to be
methods are effectively similar to paper watermarks, superior compared to spatial-domain method . It is
creating markings in an image. This can be more imperceptible and robust though more
achieved, for example, by modifying the luminance complex. With the advent of high speed Internet and
of parts of the image. While masking does change demand for larger payload, video signal will be the
the visible properties of an image, it can be done in perfect cover signal for the years to come.
such a way that the human eye will not notice the
difference. Since masking uses visible aspects of the III. PROPOSED METHOD
image, it is more robust than LSB modification with There are two distinct method to embed a
respect to compression, cropping and different kinds secret message file in the cover file. The block
of image processing. The information is not hidden diagram of video steganography (Encoding ) is
at the noise level but is in the visible part of the shown in fig2.
image which makes it more suitable than LSB

1642 | P a g e

(1) We encrypt the secret message file using simple video itself. With the help of this index, the frames
bit shifting and XOR operation in the secret containing the secret information are located.
message file. Hence, during the extraction process, instead of
(2) The encrypted secret message we embed in the analyzing the entire video, the frames containing the
cover file in alternate byte. We substitute bits in secret data are analyzed with the help of index at the
LSB and LSB+3 bits in the cover file. receiving end. When steganographed by this
method, the probability of finding the hidden
1. Bit Exchange method information by an attacker is lesser when compared
Simple bit exchange method is introduced for to the normal method of hiding information frame-
encrypting any file. The following are the steps for by-frame in a sequential
encryption method.
 Step-1: Read one by one byte from the secret Cover
message file and convert each byte to 8-bits Video file
Then we apply 1 bit right shift operation on the
entire file so that each byte will be modified
accordingly.
 Step-2: We read 8 bits at a time and divide into Collect
two blocks 4 bits each and then perform the information of
XOR operations with 4-bits on the left side with video
4 bits on the right side and substitute the new
bits in right 4-bit positions. The same thing
repeated for all bytes in the file.
 Step-3: Repeat step-1 by performing 2 bits right
Cover Frames
shift for all bytes in the secret message file.
Then repeat step-2 again.
In this paper, maximum 5 bits are shifted to
right. But the user can take more bits for right shift
operation to make the secret file more random. In
the decryption process we follow the reverse
process. Secret Frame
message Other
2. Steganography Algorithm:
In this paper we have use the substitution frames
of LSB and LSB+3 bits of the cover file in alternate
bytes. The last 300 bytes of the cover file we use for Stego frame
embedding password, size of the secret message file.
After that we start to embed the secret message file.
We read one byte from encrypted secret message
file and convert it into 8 bits and then we take 2 bits
of the encrypted secret message and substitute the Stego frames
LSB and LSB+3 bits of the cover file and then leave
one byte of the cover file intact. Then again
substitute 2 bits. The same process we repeat for all
8 bits of the secret message. Fig 2. Block diagram of LSB Video Steganography
Here the change is prominent as we embed technique (Encoding). manner. It also reduces the
in text characters but if we do the same in some computational time taken for the extraction process.
image then the changes made here will not be very
significant as our eye will not be able to differentiate 4. Stego Key
between two colors. To embed secret message we To make the system more secured we have
have to first skip 300 bytes from the last byte of the introduced the password(Stego key) while
cover file. After that we start to embed bits of the embedding an encrypted secret message file. If
encrypted secret message into the cover file. The password is correct then the program will read the
size of the secret message file must be less than 10% file size from the cover file and start to work on the
of the cover file. For .EXE or .DOC file the size of cover file.
the secret message file must be 1-5% of the cover To extract the secret message we perform
file. exactly the reverse process of the encryption
method. The program first matches the password. If
3. Index Creation it is correct then it will read the size of the secret
We can also create index for the secret message file from the embedded cover file. Then it
information and the index is placed in a frame of the will read 8 bytes and extract 8 bits from 4 alternate

1643 | P a g e

bytes and convert them to a character and write onto
an external file. Once all bytes extracted from the REFERENCES
cover file then we run the decryption program to get [1] A.Nath, S.Das, A.Chakrabarti, Data Hiding
the original secret message file. This steganography and Retrieval, Proceedings of
algorithm can be applied on different cover files IEEEInternational conference on
such as image file, audio file, video file, word Computer Intelligence and Computer
file, Excel file, Power point file, .exe file. This Network held at Bhopal from 26-28 Nov,
method could be most appropriate for hiding 2010.
any file in any standard or non standard cover [2] Samir K Bandyopadhyay, Debnath
file such as word, excel, .ppt, .exe, image, Bhattacharyya1, A Tutorial Review on
audio, video files. Steganography, UFL & JIITU, IC-2008,
pp. 105-114.
5. ADVANTAGES [3] A.Nath, S.Ghosh, M.A.Mallik, Symmetric
 Highly secure key cryptography using random key
Since random data are also placed in generator, Proceedings of International
unused frames in the video, the attacker is left conference on SAM- 2010 held at Las
clueless to know the real secret data hidden in the Vegas(USA) 12-15 July, 2010, Vol-2, pp.
video. Hence highly confidential data like military 239-244
secrets and bank account details can be easily [4] Arup Kumar Bhaumik, Minkyu Choi,
steganographed in ordinary video and can be Rosslin J. Robles, and Maricel O.
transmitted over internet even in unsecured Balitanas, Data Hiding in Video,
connection. International Journal of Database Theory
 Capacity and Application, June 2009.
Text based steganography has limited [5] Joyshree Nath, Sankar Das, Shalabh
capacity and Image steganography tried to improve Agarwal and Asoke Nath , Advanced
the capacity where 50% of original image size can steganographic approach for hiding
be used to hide the secret message. But there is encrypted secret message in LSB, LSB+1,
limitation on how much information can be hidden LSB+2 and LSB+3 bits in non-standard
into an image. Video Steganography has been found cover files, International Journal of
to overcome this problem. Computer Applications(0975-8887) Vol 14-
 Imperceptibility No7, Feb 2011
Lowest chances of perceptibility because of [6] Yongjian Hu, and Heung Kyu Lee, and
quickly displaying of the frames, so it’s become Jianwei Lee, DE-based reversible data
harder to be suspected by human vision system. hiding with improved overflow location
 Video error correction map, IEEE Transaction on circuits and
Since the transmission of any data is always systems for video technology, volume 19,
subject to corruption due to errors, then the video number 2, February, pp. 250-260, 2009.
transmission must deal with these errors without [7] Joyshree Nath and Asoke Nath, Advanced
retransmission of corrupted data. This is another Steganography Algorithm using encrypted
application for steganography rather than security secret message, International Journal of
purpose. Advanced ComputerScience and
 Less computational time Application (IJACSA)Vol-2 No.3, pp. 19-
Since use of indexing concept, the process of 24, March (2011).
retrieving the secret data from the steganographed [8] Agniswar Dutta, Abhirup Kumar Sen,
video becomes very simple and requires very less Sankar Das, Shalabh Agarwal, Asoke Nath,
time. New Data Hiding Algorithm in MATLAB
using Encrypted secret message,IEEE
IV. CONCLUSION ICCSNT-201,pp.262-267.
In this paper a robust method of [9] R. Balaji, G. Naveen, Secure Data
imperceptible audio, video, text and image hiding is Transmission Using Video Steganography.
proposed. This system is to provide a good, efficient [10] V.S.,K.Balasuhramaniyam ,N.Murali,
method for hiding the data from hackers and sent to M.Rajakumaran, Vigneswari, Data Hiding
the destination in a safe manner. Thus it can be in Audio Signal, Video Signal, Text and
concluded that Bit Exchange Method and indexing JPEG Images,IEEE ICAESM -2012, pp.
in proposed method can be used which will bring 741-746.
various advantages which can be used for a number (11) G. Doerr and J. Dugelay, "Security pitfalls
of purposes other than covered communication or of frame-by-frame approaches to video
deniable data storage. watermarking", IEEE Transactions on
Signal Processing, vol. 52, 2004, pp. 2955-
2967

1644 | P a g e

It3116411644

More Related Content

What's hot (19)

Similar to It3116411644 (20)

It3116411644