An Energy Efficient and High Speed Image Compression System Using Stationary Wavelet Transform

International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 8 240 – 243
_______________________________________________________________________________________________
240
IJRITCC | August 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
An Energy Efficient and High Speed Image Compression System Using
Stationary Wavelet Transform
V Srinivasa Rao
Dept of ECE
SVECW
Bhimavaram, AP, India
Rajesh K Panakala
Dept of ECE
PVPSIT
Vijayawada, AP, India
P.Rajesh Kumar
Dept of ECE
AUCE
Vizag, AP, India
Abstract— Image compression is one of the interesting domains nowadays in all areas of research. Everybody working with huge of amount of
data in their daily life. In-order to deal with such huge amount of data, there is a need to store and compress the data. So there is a need to
develop a system to compress and store the data. JPEG 2000 is a system to achieve this object. In this paper an area efficient and high speed
JPEG2000 architecture has been developed to compress the image data. To implement JPEG2000 system, here a transform called stationary
wavelet transform has been used. Stationary wavelet transform reduces the bottlenecks existing in the wavelet transform. Stationary wavelet
transform avoids the problem of invariance-translation of the already existing discrete wavelet transform. The proposed stationary wavelet
transform based JPEG2000 improves the speed and efficiency of power compared to the discrete wavelet transform based JPEG2000. Many
image compression applications such as tele-medicine, satellite imaging, medical imaging require high-speed, low power compression
techniques with small chip area. This paper has an analysis on the speed of JPEG2000 using stationary wavelet transform and it will be
compared theoretically and practically with the JPEG2000 using discrete wavelet transform. The amount of information missing in the test
image usually been very small when compared to the DWT based JPEG2000.The MSE and PSNR values proved to be better when compared to
the DWT based JPEG2000. The proposed SWT based JPEG2000 compresses and decompresses the image at a faster rate than the DWT based
JPEG2000.Finally the design will be implemented in XILINX Virtex-4 FPGA Kit. .The power consumption of the proposed method proved to
be 290mW compared to other types of compression techniques.
Keywords—Compression, Stationary wavelet transform, Discrete wavelet transform JPEG2000, MSE,PSNR.
__________________________________________________*****_________________________________________________
I. INTRODUCTION
JPEG2000 is an efficient image compression technique
when compared to the JPEG. Image compression plays a
vital role in many applications such as tele medicine,
satellite images, internet etc. Already existing architectures
of JPEG2000 consumes more power and compresses the
image at a slower rate. Hence there is a need to develop an
architecture which consumes less energy and operates at
higher speed. The conventional DWT used in image
compression is not shift variant. This means that the DWT
translated version of a signal is not same as the original
signal. Hence there is a need to use stationary wavelet
transform based JPEG2000 image compression system. In
stationary wavelet transform(SWT) the high pass filters and
low pass filters are applied to the data in the block segments
of an image. In SWT modify the filters by padding zeroes.
Stationary wavelet transform based JPEG 2000 is slightly
computational intensive than discrete wavelet transform
based JPEG2000[1][2]. In multi-scale signal processing,
wavelet is a time-frequency analysis that has been widely
used in the field of image processing such as denoising,
compression, and segmentation. For each modification in
the circuit the delay and power will be reduced[3][5]. The
SWT algorithm isvery simple and close to DWT. To
calculate the decimated DWT for a given signal of length by
computing approximation and detail coefficients for every
possible sequence. The simulation results show the
reduction in power and delay. The stationary wavelet
decomposition is more tractable than the wavelets.SWT has
the advantage of maintaing the same number of coefficients
throughout all scales. SWT having 2nk coefficients where n
is the length of the signal and k is the number of scales is
having high redundancy which is particularly suitable for
image compression applications.
The paper is organized as follows: Section I deals with
introduction, section II deals with related work, section III
covers proposed work, section IV covers results and section
V states the conclusion of the work.
II. REVIEW OF PREVIOUS WORK
A) Stationary Wavelet Transform (SWT)
Among the different tools of multi-scale signal processing,
wavelet is a time-frequency analysis that has been widely
used in the field of image processing such as denoising,
compression, and segmentation. Wavelet-based denoising
provides multi-scale treatment of noise, down-sampling of
sub-band images during decomposition and the thresholding
of wavelet coefficients may cause edge distortion and
artifacts in the reconstructed images. To improve the
limitation of the traditional wavelet transform, a multi-layer
stationary wavelet transform (SWT) was adopted in this
paper, as illustrated in Figure 1.
In Figure 1, Hj and Lj represent high-pass and low-pass
filters at scale j, resulting from the interleaved zero padding
of filters Hj-1 and Lj-1 (j>1). LL0 is the original image and

Volume: 5 Issue: 8 240 – 243
_______________________________________________________________________________________________
241
_______________________________________________________________________________________
the output of scale j, LLj, would be the input of scale j+1.
LLj+1 denotes the low-frequency (LF) estimation after the
stationary wavelet decomposition, while LHj+1, HLj+1 and
HHj+1 denote the high frequency (HF) detailed information
along the horizontal, vertical and diagonal directions,
respectively[4], [5].
Fig. 1. Schematic Diagram of 2D-SWT
DWT is not a shift invariant transform. Where a SWT is a
shift invariant transform. Shift invariance is very much
important for applications such as pattern recognition, image
denoising etc.In SWT applying low pass and high pass
filters at each level but the coefficients are not decimated.
Instead of decimation padding zeroes to each coefficient.
Using a Trous Algorithm,upsampling the filter coefficients
by inserting zeros shown in figure 1.
Fig. 2 Block diagram of a stationary wavelet transform

Volume: 5 Issue: 8 240 – 243
_______________________________________________________________________________________________
242
_______________________________________________________________________________________
Sub-band images would have the same size as that of the
original image because no down-sampling is performed
during the wavelet transform. In this study, the Haar wavelet
was applied to perform multi-layer stationary wavelet
transform on a 2D image [6].
Mathematically, the wavelet decomposition is defined as:
LLj+1(χ, γ) = L[n] L [m] LLj (2j+1 m- χ, 2j+1 n- γ)
LHj+1(χ, γ) = L[n] H [m] LLj (2j+1 m- χ, 2j+1 n- γ)
HLj+1(χ, γ) = H[n] L [m] LLj (2j+1 m- χ, 2j+1 n- γ)
HHj+1(χ, γ) = H[n] H [m] LLj (2j+1 m- χ, 2j+1 n- γ)
Where L[·] and H[·] represent the low pass and high pass
filters respectively, and LL0(X,Y)=F(X,Y)
III. PROPOSED WORK
A) COMPRESSION
THE FOLLOWING STEPS SUMMARIZE THE PROPOSED METHOD:
1. Load the individual blocks of an image in a order.
2. Perform a stationary wavelet decomposition of the
elements in a block. SWT is implemented to obtain non
decimated-wavelet coefficients.
3. Construct approximations coefficients and details
coefficients from the previous step.
4. Show the approximation and detail coefficients at level 1.
5. Successive separation of the coefficients and up sampling
of the decomposition filters is repeated until the best tree
structure at a predefined decomposition level is obtained.
6. For each wavelet decomposition level , a level dependent
soft threshold is determined.
7. Soft thresholding is applied on the wavelet coefficients.
8. The reconstructed image can be obtained from the
approximation and detail coefficients
Fig.3.Architecture of proposed algorithm
SWT offers better denoising and compression
capability than the existing wavelet transforms. The main
advantage of SWT is it is shift invariant transform. The
SWT algorithm is very close to DWT but in SWT the down
sampling operation after filter convolution is suppressed.
The division obtained is then a redundant representation of
the input data. The advantage of this redundant
representation over the memory-efficient decimated DWT is
the reduction of problems at discontinuities and
irregularities in reconstructed image. These problems are
caused by unpredictable changes in coefficients with
different time shifts. Preprocessing phase receives images as
input, so that the proposed approach resize the image in
accordance with the measured rate of different sizes to (8 ×
8) and then converted from RGB to gray scale.
B) FPGA Implementation
The flexibility offered by FPGA makes the FPGA is
suitable for modifications in the design. Moreover FPGA
offers less consumption, less cost, high speed and high
reliability. In this work image compression system has been
implemented by using stationary wavelet transform. The
newely developed methods for image compression system
has been mentioned in[4],[5].The system has been
implemented in software using the soft processor Vertex
from Xilinx, which facilitates interaction with peripherals.
IV. RESULTS
MATLAB simulations have been run to test the proposed
algorithm. The performance of the compression algorithm is
assessed in terms of compression ratio, mean square error
etc.
Table 1 shows the comparison of results among various
types of compression techniques. The proposed work proved
to be
Better compression time and compression ratio compared to
other compression techniques.
Table 1: Comparison of Results
Compression
Technique
Image Compression
time(Sec)
Compression
ratio
2D-SWT Cameraman 0.0315 4.35
2D-DWT Lena 0.412 5.12
2D-DCT Peppers 1.21 5.74
Table 2 shows that the synthesis report of the proposed
method has been implemented on Vertex 4 FPGA device.
The proposed architecture consumes less power and
operates at higher speed.

Volume: 5 Issue: 8 240 – 243
_______________________________________________________________________________________________
243
_______________________________________________________________________________________
Table 2: Synthesis report
Hardware
utilization
Used Available % of
utilization
LUTs 125 1214 10.29
Muxs 252 2560 9.84
IOBS 122 520 23.46
Power
Consumption
290mW --- ----
Delay 20ns ---- -----
V. CONCLUSION
The image compression system using stationary wavelet
transform has been proved to be an efficient method for
compressing and decompressing an image than Discrete
wavelet transform based image compression system. The
compression system has been verified using reconfigurable
field programmable gate arrays to achieve better
performance at low bit rates. The power consumption
proved to be less than the existing methods using discrete
wavelet transform. The system compresses and
decompresses the image at a faster rate with less critical
path delay. The proposed image compression system
performs with more accuracy. The reconstructed image is
appears to be same as the input image by losing few
redundant data. If this concept is applied to 3-D stationary
wavelet transform it will give some interesting results in
image compression systems.
REFERENCES
[1] Kaur.M,Kaur.G , “A survey of lossless and lossy image
compression technique”,International advanced research
computaion and communication engineering ,Vol.3, No.2,
Feb. 2014.
[2] Taubman DS,Marcelin MW, “JPEG2000 image
compression fundamentals ,standards and practices”,
Springer, 2002.
[3] Gonjalez R , “Digital image compression using MATlab,”
Pearson education,March 2004.
[4] GP Nason and BW Silverman, “Stationary Wavelet
Transform and some Statistical Applications ,” in Proc.
IEEE Int. Midwest Symp. Circuits Syst. Dig. Tech. Papers,
Aug. 2010, pp. 893–896.
[5] Omar Ghazi Abbood , Mahmood A. Mahmood ,“Hybrid
Compression based Stationary Wavelet Transforms ,” in
International journal of Engineering,managemnet and
technology , Nov 2016, pp. 113–118.
[6] Md.Ghaji, “Lossy Compression Using Stationary Wavelet
Transform and vector quantization” IEEE Image
compression Electron. Lett., vol. 39, no. 12, pp. 894–895,
Jan. 2013.
[7] S. Porwal ,Chowdary Y ,Joshi J, “Data compression
methodologies for lossless data and comparision between
algorithms,”I J Eng Sci Innovative Technology, vol. 7, no.
8, pp. 1680–1687, Aug. 2003.
[8] Sathappan S. A vector quantization technique for image
compression using modified fuzzy possibilistic C-means
with weighted mahalanobis distance. Int J Innovative Res
Comput Commun Eng 2013; 1(1): 12-20. 316–317.
[9] Dwivedi A, Bose NS, Kumar A, Kandula P, Mishra D,
Kalra PK, “ A novel hybrid image compression technique”,
Proceedings of the ASID 6, 8-12 Oct, 2012, New Delhi;
2012. pp. 492-5.
[10] Senthi Shanmugasundaram and L.Robert, “A Comparative
Study Of Text Compression Algorithms”,International
journal of communications and technology,December
,2011.
[11] Nema M, Gupta L, Trivedi NR. Video compression using
SPIHT and SWT wavelet. Int J Electron Commun Eng
2012; 5(1): 1-8 .

An Energy Efficient and High Speed Image Compression System Using Stationary Wavelet Transform

More Related Content

What's hot (17)

Similar to An Energy Efficient and High Speed Image Compression System Using Stationary Wavelet Transform (20)

More from rahulmonikasharma (20)

Recently uploaded (20)

An Energy Efficient and High Speed Image Compression System Using Stationary Wavelet Transform