steganography based image compression

International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-2, Issue-6, June- 2016]
Infogain Publication (Infogainpublication.com) ISSN : 2454-1311
www.ijaems.com Page | 685
Steganography based image compression
Rajandeep Kaur1
, Pooja2
1
Department of Computer Science, CTIEMT, IKG PTU, Jalandhar, Punjab, India
2
Department of Computer Science, CTIEMT, Jalandhar, Punjab, India
Abstract— The intention of image compression is to discard
worthless data from image so as to shrink the quantity of data
bits favored for image depiction, to lessen the storage space,
broadcast bandwidth and time. Likewise, data hiding
convenes scenarios by implanting the unfamiliar data into a
picture in invisibility manner. The review offers, a method of
image compression approaches by using DWT transform
employing steganography scheme together in combination of
SPIHT to compress an image.
Keywords— DCT (Discrete Cosine Transform), DWT
(Discrete Wavelet Transform), CR (Compression Ratio),
PSNR (Peak Signal-to-Noise Ratio).
I. INTRODUCTION
The main inspiration at the back image of compression is to
reduce the surplus and irrelevance data from the picture, to
achieve the aim of placing it away and exchanged skillfully.
As a result, condensed picture is symbolized by smaller bits
count in comparison to actual. Therefore, the compulsory
storage space decreased; as a consequence huge quantity of
pictures can be stocked up into the same storage space and
stimulated in faster way to standby the broadcast bandwidth,
and time. In picture compactness/compression technique, for
nearly all part supernatural and spatial extra data should be
reduced as a large amount as could realistically be predicted.
Picture compactness is utilized in many applications to
efficiently enhance effectiveness and performance.
Due to this reason many compactness/compression
techniques i.e. scalar or vector quantization, differential
encoding, predictive, and Transform picture coding have
been shown. At small bit rate Transform coding is efficient
among these methods. [1]. Transform coding relies on the
regulation that pixel in a picture show an exact height of
association with their neighboring pixels. Consequently, these
associations can be oppressed for the judgment of a pixel
height from its appropriate adjacent pixel. An alteration in
this way shows to drawing of correlated (spatial) data into
uncorrelated (transformed) coefficients. Therefore, alteration
have to employ the means that the information of every pixel
is generally extremely minute i.e., to a great degree
illustration fraction of a pixel be able to be considered by
making utilize of its adjacent pixels. Depending upon the
compactness technique used, the picture can be regenerated
with and with no visual failure. In lossless compactness
method, the regenerated picture after compactness is
numerically impossible to tell apart from the actual. While, in
lossy compactness/compression method, the regenerated
picture comprised of poverty in deference to the actual, as
these methods source picture excellence deprivation in every
compactness/compression or decompression step.
Consequently, every lossy methods used for picture
compactness/compression having great
compactness/compression rate in comparison to lossless
techniques i.e. fine class of compressed/compacted picture
with a fewer quantity of compactness, whereas lossy-
compactness/compression methods [2] direct to data failure
with elevated compactness/compression ratio. Lossy
compactness/compression methods comprise of predictive
and transform coding. Transform coding method chiefly
concerns a Fourier transform; i.e. Discrete Cosine Transform
and DWT Transforms [3]. In the presented study, a picture
compression approach using a hybrid method of DWT and
SPHIT in combination with steganography is employed.
The rest paper is planned as track: Section 2 put in plain
words Compression Fundamental Techniques; Section 3
elaborates data hiding concepts; Section 4 incorporated
proposed methodology and last Section 5 provides the
conclusions.
II. COMPRESSION FUNDAMENTAL
TECHNIQUES
The essential backdrop is offered here in this segment.
Segment 2.1 and 2.2 momentarily elaborates the lossy JPEG
compactness/compression technique based on DWT and
DCT.
2.1 Joint Photographic Expert Group (JPEG)
JPEG stands for joint photographic expert group and it is a
lossy compression based format and also known as ISO/ITO-
T and created in 1980.The basic JPEG forms generally of the
subsequent steps. The all three RGB layers of the novel color
picture are transformed to YCBCR. Then each layers of
YCBCR is divided into chunk of dimension 8 X 8. The

subsequent operations are applied on every chunk: DCT
(discrete cosine transform), quantization, crisscross scan,
RLE (run length encoding) and Huffman coding. The
Discrete Cosine Transform draws the chunk of 8 X 8 pixels
into a group of altered coefficients. Compression is achieved
through quantization, by dropping the quantity of bits used to
represent every coefficient. The coefficients are then encoded
in a crisscross manner from small frequency coefficient to
elevated frequency coefficients. The AC coefficients are
encoded by Huffman coding and RLE. The DC coefficient
generally has a importance to great extent than AC
coefficients, and there is a extremely close association
between the DC coefficients of neighboring chunks.
Consequently, JPEG technique encodes the distinction
between the DC coefficients of adjacent chunks using DPCM
(Differential pulse code modulation) encoding.jpeg
compression helps to reduce the image size without
degrading the image quality of image. Whereas JPEG
pictures can have bright, high-resolution picture information,
it is a lossy arrangement, which indicates few superiority is
lost when the picture is compacted/compressed. If the picture
is compacted/compressed to a great extent, the graphics turn
into obviously "blocky" and a little of the feature is
misplaced. Like GIFs format, JPEG formats are fractious
platform, implicating the matching file will seem the same on
both PC and Mac. The entire cause after this JPEG
arrangement is to create lesser size workstation files.
Uncompressed pictures can be enormous information files.
By stocking the pictorial information in the JPEG
arrangement it is probable to diminish the extent of
information file. Here are two key features of JPEG that
needs to be measured before storing a file using the
arrangement. Both narrate to the "lossy" character of JPEG it
really modify the picture and eliminates a quantity of the
unique picture information. How this influence the picture is
prohibited by two chief features: the picture type and
compactness/compression amount. JPEG images has high
quality and used for compression process. It supports to lossy
compression. The pixels resolution of JPEG images very high
and clear to visual due to JPEG compression its pixels
resolution destroy and harsh to visible and does’t support
clearly to logo based images.
2.2 JPEG2000
The major dissimilarity between JPEG2000 and JPEG is in
the use of Discrete Wavelet Transform instead of Discrete
Cosine Transform. The JPEG2000 forms generally of the
subsequent steps. The all three RGB layers of the novel color
picture are transformed to YCBCR. Then each layers of
YCBCR is divided into minor non overlapping chunks, known
as strips. All processes, like Discrete Wavelet Transform
(DWT), quantization and entropy encoding are completed
separately on every strip. The strip may be as huge as the
whole picture dimension or lesser separations. The Discrete
Wavelet Transform is processed by sorting every line and
piece of the picture strip with a elevated pass and low pass
filter. Daubechies filters and normally DB9/DB7 is used
while applying wavelet transform. The yield from every filter
is further down sampled by 2 so that the sample rate stays
steady. As a consequence of application of DWT, the strip is
separated into four sub teams LL, HL, LH, and HH that holds
coefficients with completely different straight and
perpendicular spacial frequency characteristics. JPEG 2000
offers each lossless and lossy compression within the
matching file flow, whereas JPEG typically solely uses lossy
compactness/compression. JPEG will contain a lossless
compactness/compression locomotive; however it's broken
free the lossy engine, and isn't used fairly often. Thus, once
prime superiority could be a concern, JPEG 2000 proves to
be a far higher compression tool. JPEG 2000 guarantees the
next quality final image, even once victimization lossy
compression. Since the JPEG 2000 format includes abundant
richer content than existing JPEG files, rock bottom line
result is that the ability to deliver abundant smaller files that
also contains a similar level of detail because the larger
original JPEG files. 256 channels of data can be handled by
the JPEG 2000 files in comparison to the JPEG standard,
which, by motive of ordinary accomplishment, is restricted to
only red, green, and blue channels data. The Joint
Photographic Experts Group 2000 (JPEG 2000) customary,
completed in 2001, describes a new picture coding method
using high-tech compactness/compression scheme based on
wavelet skill. Its structural design is functional for a lot of
varied functions, together with Internet picture sharing,
defense organizations, digital cinematography, and
therapeutic imaging. JPEG 2000 is its skill to show pictures
at unusual declarations and dimensions from the similar
picture file. With JPEG, a picture file was merely talented to
be shown a sole method, with a convinced decision. As
JPEG 2000 is supported on wavelets, the wavelet flow can be
only incompletely decompressed if the consumer merely
desires a small decree picture, whilst the complete decree
picture can be examined if this is need.
III. DATA HIDING USING STEGANOGRAPHY
Steganography is that the ability of concealment covert
communication within the mass medium with the intend of
no one will presume the existence of the message. Associate
in Nursing everyday methodology for implanting a
knowledge (payload) within the media lacking any sensory
activity deformation within the media is to control the

smallest amount vital little bit of the image pixels and for that
reason represented as least vital bit (LSB) steganography. In
JPEG pictures, covert knowledge bits square measure hid into
the Least Significant Bit (LSB) of smoothed quantal DCT
coefficients; apart from coefficients values (-1, 0 or 1). The
necessities for the judgment of coefficients think about
keeping far from possible ambiguity within the secret
knowledge unraveling practice. For design, if the covert
knowledge bit is zero and DCT constant price is one,
therefore constant price are going to be modified to zero.
Meanwhile, dissimilar coefficients with the initial esteem
zero do not have any anonymous info constituted in them.
during this manner, numerous coefficients don't seem to be
used for implanting. Compactness/compression exploitation
steganography ought to accomplish constraints i.e. data size,
robustness, and physical property. The hide bits square
measure in cognizable, if an individual with average vision
ineffectual to spot media that embrace hid knowledge from
the persons that do not have. Added, data capacity refers to
the entire bits obscured in media, whereas satisfying the
physical property limitation. The constituted knowledge is
powerful on the off probability that it may be recognized
once non intentional adjustment, as an example, lossy
compression. The final necessity to that deliberation should
be paid within the wake of implanting method is file size and
this necessity isn't monumental in typical steganography
reason, but once compactness/compression the file size ought
to be compressed. Knowledge concealment methods have in
used vital function with the quick climb of rigorous relocate
of transmission covert and content messages. Steganography
is that the skill of concealment info in conducts in which
forestall discovery. The Steganography used to transport info
from one place to different place through public channel in
covert manner. Steganography conceals the terribly
subsistence of a data so if creative it typically attracts no
suspicion the least bit. Steganography means that
concealment a covert message (the implanted message) inside
a bigger one (base wrap) in such some way that Associate in
Nursing spectator cannot notice the attendance of substances
of the concealed data. A lot of alternative transporter file
arrangements may be used; however digital pictures square
measure the foremost well-liked due to their frequency on the
net. For concealment secret info in pictures, there exists an
oversized type of Steganography methods a number of
methods are supplementary composite than rest and all
contain own muscular and scrawny points. Dissimilar
functions have different needs of the Steganography method
used. Various techniques of hiding data using steganography
are:
• LSB (Least Significant Bits)
• DCT (Discrete cosine Transform)
• DWT (Discrete wavelet transform)
3.1 LSB (Least Significant Bit) Technique
The most popular technique in steganography is LSB. It is
utilized to conceal the covert data bits in both grey scale and
colored images based on its binary coding. LSB algorithm is
used to hide the secret message by replacing the right-most
two LSBs of the pixels. LSB makes the change in the
brightness of the image and it is easy for the intruders to
attack on the image. The thought of Least Significant Bit
implanting is easy. It develops the reality that the height of
accuracy in a lot of picture patterns is extreme better than that
perceivable by normal individual visualization.
Consequently, a tainted picture with small disparity in its
shades will be impossible to differentiate from the actual by
an individual, immediately by looking at it. LSB method can
be 1bit or 2 bit. Figure 1 represents 2 bit LSB.
Fig.1:Explanation of LSB Technique
IV. PROPOSED METHODOLOGY
Steganography based compression techniques must convince
imperceptibility, data capacity, and file amount needs. The
implanted statistics is imperceptible if a person subject with
usual visualization is not capable to differentiate medium that
hold secreted statistics from those that do not. The data
capacity means the amount of secreted bits simultaneously
pleasing the invisibility need. Lastly, the final necessity to
which notice have to be compensated post implanting is file
size. This need is not important in the common data hiding
functions, but for the compactness/compression it has to be
condensed. Two processes of compression are presented: one
is compression using DCT and steganography, the other is
compression using DWT and Steganography. The
methodology of implementing both of the algorithms is
explained in Fig. 2.
1 0 1 0 0 0 1 1
0011
Insert two bits directly
Image Pixel

Fig.2:Explanation of Proposed Methodology
Phase1. Initial Processing: In this phase, the input image is
loaded of jpeg format and then decomposes it into RGB
layers. RGB layers means to an organization for in place of
the shades to be utilized on a PC screen. All these three
colors (RGB) can be joined in a variety of magnitudes to gain
one shade in the observable spectrum. Heights of Red, Green,
and Blue can vary with in 100 percent of complete strength.
Every height is symbolized by the choice of decimal
numerals ranging from 0 to 255 (every color contains 256
heights), corresponding to the variety of binary numerals
from 00000000 to 11111111, or hexadecimal numerals 00 to
FF. The entire count of obtainable colors is 256 x 256 x 256,
or 16,777,216 probable shades.
Phase2. DWT Implementation: In this phase, DWT is
concerned on input picture and the selected picture is divided
into four blocks i.e. LL, LH, HL, HH and obtained
manipulated coefficients Discrete Wavelet Transform, the
key in information is overtaken throughout set of mutually
high pass and low pass filter in both directions, through
columns and rows. The outcomes are then down sampled by
2 in every way as in case of 1 Dimensional Discrete Wavelet
Transform (DWT) [6]. Fig. 3 shows, outcome is attained in
group of four coefficients LL, HL, LH and HH. The initial
alphabet symbolizes the change in line where as the next
alphabet symbolizes change in column. The alphabet L in
coefficients implies low pass signal and alphabet H implies
high pass signal. LH coefficients are obtained by applying
low pass filter in row and a high pass filter along column.
Fig. 3:Discrete Wavelet Transform
Phase 3. Quantization: In Quantization part, the high
frequency values that area unit obtained by applying DWT
area unit separated from low frequency bands and
quantization can convert the getting floating values into
number quantization makes the vary of a symbol distinct, so
the quanta signal takes on solely a distinct, typically finite, set
of values in contrast to sampling (where we saw that
underneath appropriate conditions precise reconstruction is
possible), quantization is typically irreversible and ends up in
loss of data. It therefore introduces distortion into the quantal
signal that can't be eliminated. One in all the essential
decisions in quantization is that the range of distinct
quantization levels to use. The basic trade-off during this
selection is that the ensuing signal quality versus the quantity
of knowledge required representing every sample. Basically
it's an offer a around off price like one.222 into modification
one. Quantization suggests that floating price grow to be the
integer price.
Phase4. SPIHT: It suggests that SET PARTITION IN
HIRARICAL part. This method set the low band values
within the variety of binary tree. This method helps to
enhance the standard of image and use less memory. The
SPIHT technique isn't an easy extension of ancient strategies
for compression, and represents a crucial advance within the
field. Here, we tend to area unit adding low frequency
coefficients and sent it to SPIHT and it represents these
values into a hieratical tree. One in all the foremost
economical algorithms within the space of
compactness/compression is that the group Partitioning in
hierarchic Trees (SPIHT). In essence it uses a sub-band
technologist; to provide a pyramid arrangement wherever a
picture is rotten consecutive by affecting power
complementary high pass and low pass filters and so
decimating the ensuing pictures. This area unit one-

dimensional filters that area unit applied in cascade (row then
column) to a picture whereby making four-way
decomposition: LL (low-pass then another low pass), LH
(low pass then high pass), hectoliter (high and low pass) and
eventually HH (high pass then another high pass). If we tend
to do modification in low filter then it will be simply
establish that we tend to area unit victimization high
frequency or not.
Phase5. Steganography: In this phase, data hiding is used to
embed the hide the data behind any type of media file and
here we are using BMP image to embed the coefficients using
LSB Technique. In this, the least significant bits are replaced
with coefficient values and image is compressed. It hide the
data, the unauthorized person cannot access the data so these
techniques are used in image compression. Fundamentally,
the data implanting procedure in a steganographic
organization founds by recognizing a wrap medium’s
superfluous bits (those that can be modified without
destroying that medium’s integrity). The implanting
procedure generates a stego medium by restoring these
superfluous bits with information from the concealed
message.
V. CONCLUSIONS
In this study a new Image compression approache for images
is prepared based on DWT, steganography, and SPIHT. This
technique will enhance the performance of the conventional
JPEG and JPEG2000 techniques of compression. Here
discussed technique, DWT may offer improved compression
ratio with no loss of additional data of image. In wavelet,
there is no convincing reason to chunk the image. It
facilitates energetic transmission of the image. As outcomes
proves that DWT offers added compression in contrast to
conventional JPEG compression using DCT. Also DWT give
further storage saving and fine quality of regenerated image.
It is further appropriate for steady applications as it is having
an upright compression ratio together with safeguarding the
majority of the information.
REFERENCES
[1] Muhammad AzharIqbal, Dr Muhammad YounusJaved,
UsmanQayyum, “Curvelet-based Image Compression
with SPIHT”, IEEE computer society. International
Conference on ConvergenceInformation Technology,
2007.
[2] M. Ghanbari "Standard Codecs: Image Compression to
Advanced Video Coding" Institution Electrical
Engineers, ISBN: 0852967101, 2003.
[3] Chuanwei Sun 1Quanbin Li,Jingao Liu “The Study of
Digital Image Compression Based on Wavelets”,
ICALIP2010.
[4] Kencabeen and Peter Gent,"Image Compression and the
Descrete Cosine Transform"Math 45,College of the
Redwoods.
[5] K. R. Rao and P. Yip, Discrete Cosine Transform:
Algorithms, Advantages, Applications. Boston:
Academic Press, 1990.
[6] Anoop Mathew and Bensiker Raja Singh D, “Image
Compression using Lifting based DWT”, International
Journal of computers, Information Technology and
Engineering, Jan-June 2009.
[7] E. Elharar, Adrian Stern, OferHadar,“A Hybrid
Compression Method for Integral ImagesUsing Discrete
Wavelet Transform andDiscrete Cosine Transform”,
Member, IEEE, and Bahram Javidi, Fellow, IEEE.
[8] Sriram.B,Thiyagarajans.S, “Hybrid Transformation
technique for image compression ”,Journal of
theoretical and applied information technology, 31st
July 2012 Vol. 41 No.2.
[9] Campisi, P., Kundur, D., Hatzinakos, D., & Neri, A.
(2002). Compressive data hiding: An unconventional
approach for improved color image coding. EURASIP
Journal on Applied Signal Processing, 2002(2), 152–
163.
[10]Chen, B., & Wornell, G. W. (2001). Quantization index
modulation: A class of provably good methods for
digital watermarking and information embedding. IEEE
Transaction on Information Theory, 47(4), 1423–1443.
[11]Christopoulos, C., Skodras, A., & Ebrahimi, T. (2000).
The JPEG2000 still image coding system: an overview.
IEEE Transaction on Consumer Electronics, 46(4).
[12]De Queiroz, R., Choi, C. K., Huh, Y., & Rao, K. R.
(1997). Wavelet transforms in a JPEGlike image coder.
IEEE Transactions on Circuits and Systems for Video
Technology, 7(2), 419–424.
[13]Gonzalez, R. C., & Woods, R. E. (2008). Digital image
processing. Prentice Hall. Hore, A., & Ziou, D. (2010).
Image quality metric: PSNR vs. SSIM. In 20th
International conference on pattern recognition
(pp.2366–2369).
[14]Jafari, R., Ziou, D., & Mammeri, A. (2011). Increasing
compression of JPEG images using steganography. In
IEEE International Symposium on Robotic and Sensors
Environments (pp. 226–230).
[15]Solanki, K., Jacobsen, N., Madhow, U., Manjunath, B.
S., & Chandrasekaran, S. (2004). Robust image adaptive
data hiding based on erasure and error correction. IEEE
Transaction on Image Processing, 13(12), 1627–1639.

[16]Swanson, M., Zhu, B., & Tewfik, A. H. (1997). Image
coding by folding. IEEE International Conference on
Image Processing 2, 665–668.
[17]Tseng, H., & Chang, C. (2004). High capacity data
hiding in JPEG-compressed images. Informatica, 15(1),
127–142.Wallace, G. K. (1992). The JPEG still picture
compression standard. IEEE Transaction on Consumer
Electronics, 38(1), 18–35.

steganography based image compression

More Related Content

What's hot (19)

Similar to steganography based image compression (20)

Recently uploaded (20)

steganography based image compression