IRJET- An Efficient VLSI Architecture for 3D-DWT using Lifting Scheme

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 03 | Mar 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 2090
AN EFFICIENT VLSI ARCHITECTURE FOR 3D-DWT USING
LIFTING SCHEME
P. Kannan1, I. Asma2, A. Benasir3, R. Deepikha4, R. Divya5
1Professor, Dept. of ECE, Panimalar Engineering College, Poonamalle, TamilNadu, India.
2,3,4,5UG students, Dept. of ECE, Panimalar Engineering College, Poonamalle, TamilNadu, India.
----------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Image compression is one of the known
techniques in image processing. The modern real time
applications related to image processing demands high
performance discrete wavelet transform (DWT). Existing
techniques like DCT require more complex operations and
also needs increased hardware resources. Compared to
DCT, DWT has better compression results due to lossless
compression .In DWT there are two techniques ,they are
convolution and lifting scheme .The lifting scheme
represents the fastest implementation of the DWT than
the convolution based DWT. Hence lifting scheme is
considered. The proposed 3D-DWT achieves reduction in
total area and net power as compared with existing
convolution DWT and DCT. The major objective of this
work is to improve the performance of the DWT for DSP
applications.
1. INTRODUCTION
Nowadays, from satellite images to medical diagnosis
images are stored on our computers. To get these images
on the computer they must be transmitted over phone
lines or other cables. When the images are larger it takes
longer compression time and higher storage space. A
common characteristic of most images is that the
neighboring pixels are correlated and therefore contain
redundant information. Hence in order to remove that
redundant information we have to detect less correlated
pixels representation of the image. The two main
components involved in compression are redundancy and
irrelevancy reduction. In order to decrease the redundant
information we can eliminate duplication from an image
or a video. The reduction in irrelevancy eliminates the part
of the image or video that will not be noticed by the signal
receiver, namely the Human Visual System (HVS).
1.1 Redundant Techniques
Thus compression is obtained after removing one or more
of three basic data redundancies: (1) Coding redundancy,
that are present only when code words used are less than
optimal; (2) Interpixel redundancy, that occurs as a result
of correlations between the pixels of an image; (3)
psychovisual redundancy which is due to the data that is
ignored by the human visual system. For many years,
artificial neural networks (ANNs) have been studied and
used to model information processing systems based on or
inspired by biological neural structures. The artificial
neural network results with solutions whose performance
is better than that of traditional problem-solving methods,
and also provides a clear understanding of human
cognitive abilities.
Kohonen’s self-organizing map (SOM) is one of the most
popular neural network models. This is mainly introduced
for an associative memory model which is one of the
unsupervised learning algorithms with a simple structure
and computational form. Self-organization is a
fundamental pattern recognition process. In self-
organization, intrinsic inter- and intra-pattern
relationships among the stimuli and responses are learned
without the presence of a potentially biased or subjective
external influence. The SOM is mainly used to provide
topologically preserved mapping of input and output
spaces [1, 2]. The SOM is optimal for vector quantization.
The property of SOM, that provides the topographical
order mapping with enhanced fault- and noise-tolerant
abilities.
1.2 Proposed algorithm
In turn SOM is also applicable to various other
applications, like reducing dimensionality, data
visualization, clustering and classification. Many other
extensions of the SOM are also devised to extend the
mapping as an effective solution for a wide range of
applications. Wavelet transform is the only method that
provides both spatial and frequency domain information.
The properties of wavelet transform gently helps in
identification and selection of significant and non-
significant coefficients among the wavelet coefficients.
Image compression based on wavelet transform results in
an improved compression ratio as well as image quality
and thus both the significant coefficients and their
positions within an image are encoded and transmitted. In
this paper, a wavelet based image compression is applied
to the result of the SOM based vector quantization.
With the rapid progress of VLSI design technologies, many
processors based on audio and image signal processing
have been developed recently. The discrete wavelet
transform plays a major role in the JPEG-2000 image
compression standard. The Discrete Wavelet Transform
(DWT) has become a very versatile signal processing tool.
The advantage of DWT over other traditional
transformations is that it performs multi resolution
analysis of signals with localization both in time and
frequency. In addition to audio and image compression,
the DWT has important applications in many areas, such
as computer graphics, numerical analysis, radar target

distinguishing and so forth. Discrete wavelet transform
(DWT) has been widely used in many multimedia
applications including video coding and various signal
processing applications.
2. LITERATURE SURVEY
2.1 “Efficient Architectures for Two-Dimensional
Discrete Wavelet Transform Using Lifting
Scheme” by “Chengyi Xiong, Jinwen Tian, and Jian
Liu”
Novel architectures for 1-D and 2-D discrete wavelet
transform (DWT) using lifting schemes are presented in
this paper. An embedded decimation technique is
exploited to optimize the architecture for 1-D DWT, which
is designed to receive an input and generate an output
with the low- and high-frequency components of original
data being available alternately. Based on this 1-D DWT
architecture, an efficient line-based architecture for 2-D
DWT is further proposed by employing parallel and
pipeline techniques with 100 percentage hardware
utilization. Moreover, another efficient generic line-based
2-D architecture is proposed by exploiting the parallelism
among four sub band transforms in lifting-based 2-D DWT,
hence it is called high-speed architecture. The throughput
rate of the latter is increased by two times when
comparing with the former 2-D architecture, but only less
additional hardware cost is added. Compared with the
works reported in previous literature, the proposed
architectures for 2-D DWT are efficient alternatives in
tradeoff among hardware cost, throughput rate, output
latency and control complexity, etc.
2.2 “An Efficient Hardware-Based Higher Radix
Floating Point MAC Design” by “MOHAMED ASAN
BASIRI M and NOOR MAHAMMAD SK”
This article proposes an effective way of implementing a
multiply accumulate circuit (MAC) for high-speed floating
point arithmetic operations. The real-world applications
related to digital signal processing and the like demand
high-performance computation with greater accuracy.
Thus, the separate accumulation circuit can be avoided by
keeping the circuit depth still within the bounds of the
Wallace tree multiplier or Braun multiplier. In this
article, three kinds of floating point MACs are proposed.
The experimental results show improvement in worst
path delay achieved by the proposed floating point MAC
using a radix-2 Wallace structure compared with a
conventional floating point MAC without a pipeline. The
performance results show comparisons between the
proposed floating point MAC with various floating point
MAC designs for radix-2,-4,-8, and -16. The proposed
design has lesser depth than a conventional floating point
MAC as well as a lower area requirement than other ways
of floating point MAC implementation, both with or
without a pipeline.
2.3 “Area-Efficient and Power-Efficient
Architecture for High-Throughput
Implementation of Lifting 2-D DWT” by “ Basant
K. Mohanty, Senior Member, IEEE, Anurag
Mahajan, and Pramod K. Meher, Senior Member,
IEEE”(2012)
We have suggested a new data-access scheme for the
computation of lifting two-dimensional (2-D) discrete
wavelet transform (DWT) without using data
transposition. We have derived a linear systolic array
directly from the dependence graph (DG) and a 2-D
systolic array from a suitably segmented DG for parallel
and pipeline implementation of 1-D DWT. These two
systolic arrays are used as building blocks to derive the
proposed transposition-free structure for lifting 2-D DWT.
The proposed structure requires only a small on-chip
memory of words and processes a block of P samples in
every cycle, where N is the image width. Moreover, it has
small output latency of nine cycles and does not require
control signals which are commonly used in most of the
existing DWT structures.
2.4 “Image Compression Based upon Wavelet
Transform and a Statistical Threshold by “Ahmed
A. Nashat, N” and “M. Hussain Hassan” (2016)
Discrete Wavelet Transform, (DWT), is known to be one
of the best compression techniques. It provides a
mathematical way of encoding information in such a way
that it is layered according to level of detail. In this paper,
we used Haar wavelets as the basis of transformation
functions. Haar wavelet transformation is composed of a
sequence of low pass and high pass filters, known as filter
bank. The redundancy of the DWT detail coefficients is
reduced through thresholding and further through
Huffman encoding. The proposed threshold algorithm is
based upon the statistics of the DWT coefficients. The
quality of the compressed images has been evaluated
using some factors like Compression Ratio (CR) and Peak
Signal to Noise Ratio (PSNR). Experimental results
demonstrate that the proposed technique provides
sufficient higher compression ratio compared to other
compression thresholding techniques.
2.5 “Review of image compression techniques “by
“Abhipriya Singh K.G. Kirar” (2017)
Demand of multimedia growth, contributes to insufficient
bandwidth of network and memory storage device.
Therefore data compression is more required for reducing
data redundancy to save more hardware space and
transmission bandwidth. Image compression is one of the
main researches in the field of image processing. Many
techniques are given for image compression. Some of
which are discussed in this paper. This paper discusses ‘k’
means clustering, 2D-DWT and fuzzy logic based image
compression.

3. LIFTING SCHEME
This paper proposes the 3D- DWT using lifting scheme.
The lifting scheme entirely relies on the spatial domain,
has many advantages compared to filter bank structure,
such as lower area, power consumption and
computational complexity. The lifting scheme can be easily
implemented by hardware due to its significantly reduced
computations. Lifting has other advantages, such as “in-
place” computation of the DWT, integer-to-integer wavelet
transforms which are useful for lossless coding.
3.1 LIFTING ALGORITHM
In image processing, DWT can be used in image
compression, image reconstruction, image coding, and
Image fusion. In general, VLSI architecture for DWT is
classified into two categories, they are
 Convolution based and
 Lifting based.
Wim Sweldens developed a lifting scheme for the
construction of bi-orthogonal wavelets. The main
feature of the lifting scheme is that all constructions
are derived in the spatial domain. Lifting scheme is a
simple and an efficient algorithm to calculate wavelet
transforms as a sequence of lifting steps. Constructing
wavelets using lifting scheme comprises three steps:
Step 1:
 Split Samples: The original signal, input
image X (n), is split into odd and even
samples.
Step 2:
 Lifting: This step is executed as N sub steps
depending on the type of the filter, where the
odd and even samples are filtered by the
prediction and update filters, p (z) and u (z).
Step 3:
 Normalization or Scaling : After N lifting
steps, scaling coefficients K and 1/K are
applied respectively to the odd and even
samples in order to obtain the low pass sub
band i.e. significance coefficient YL (i) and
the high-pass sub band i.e. detailed
coefficient YH (i).
The proposed work is specialized for the DWT 5/3
wavelet in lifting scheme implementation. ‘X’ be the
input image, which has predefined pixels.
Let, X = [X (1), X (2) ... X (2n)] be an array of length 2n.
In this work we begin with the “poly-phase
decomposition” splitting X into two sub-bands, each
of length N. The original signal i.e. input image pixels
are split into even and odd pixels in split step of the
design.
Xo = [X (1), X (3), X (5)... X (2n-1)]
Xe = [X (2), X (4), X (6) ... X (2n)]
There are four stages in the lifting scheme
architecture which is summarized by the equations as
follows:
P1(n) = Xo(n) + a (Xe(n) + Xe(n+1))
U1(n) = Xe(n) + b ( P1(n) + Xe (n+1))
dc(n) = k * P1(n)
Sc (n) = 1/k * U1 (n)
P1 (n) and U1 (n) are scaled by the constant K and K-1
respectively, for normalizing their magnitude. Filter
coefficients are described in table 1. The inverse
transform is done by performing the lifting steps in
the reverse order and with a, b and k negated.
Fig-3.1.1: Block diagram of Lifting Scheme
Fig-3.1.2: Forward Lifting Scheme

Fig-3.1.3: Backward Lifting Scheme
4. STAGES OF COMPRESSION
4.1 INPUT IMAGE
4.2 1D-DWT
4.3 2D AND 3D-DWT
5. RESULTS
5.1 AREA-ANALYSIS
5.2 DWT LIFTING TIMING REPORT (DELAY)

5.3 MATRI X OF PIXELS BEFORE IMAGE
COMPRESSION
5.4 MATRIX OF PIXELS AFTER IMAGE
COMPRESSION
5.5 DWT-CONVOLUTIONAL OUTPUT
5.6 DWT-LIFTING OUTPUT
5.7 FPGA OUTPUT
6. ADVANTAGES
 Integer-to-integer wavelet transforms which are
useful for lossless coding.
 Better Compression Performance.
 Computation Performance is good.
 Less area and delay.
 In image processing, DWT can be used in Image
compression , Image coding, Image fusion and
Image reconstruction.
7. CONCLUSION
In this Project, Efficient VLSI architecture for convolution
based folded 1D/2D-DWTs are proposed. This project
proposes the floating point MAC based 1D/2D-DWT,

where the low/high pass FIR filter outputs are found using
a MAC and high performance VLSI architecture for
discrete wavelet transform (DWT) is proposed that are
used in real time high efficiency video coding (HEVC)
applications. The proposed 1D architecture is used to
design 2D folded and parallel designs. The performance
results show that the proposed architecture gives good
improvement as compared with existing architecture.
8. REFERENCES
[1] Po-Cheng Wu and Liang-Gee Chen, “AN EFFICIENT
ARCHITECTURE FOR TWO-DIMENSIONAL DISCRETE
WAVELET TRANSFORM”, IEEE Transactions on Circuits
and Systems for Video Technology, 2001, 11 (4), 536-545.
[2] Chu Yu and Sao-Jie Chen, “VLSI Implementation of 2-D
Discrete Wavelet Transform for Real-time Video Signal
Processing”, IEEE Transactions on Consumer Electronics,
1997, 43 (4), pp. 1270-1279.
[3] Frances comaria Marino, “Efficient High-Speed/Low-
Power Pipelined Architecture for the Direct 2-D Discrete
Wavelet Transform”, IEEE Transactions on Circuits and
Systems - II: Analog and Digital Signal Processing, 2000, 47
(12), pp. 1476-1491.
[4] Tinku Acharya and Chaitali Chakrabarti, “A Survey on
Lifting-based Discrete Wavelet Transform Architectures”,
Journal of VLSI Signal Processing, 2005, 42, pp. 321-339.
[5] Liu Hong-jin, Shao Yang, He Xing, Zhang Tie-jun, Wang
Dong-hui and Hou Chao-huan, “A Novel VLSI Architecture
for 2-D Discrete Wavelet Transform”, IEEE International
Conference on ASIC, Oct. 2007, pp.40-43.
[6] Guangming Shi, Weifeng Liu, Li Zhang, and Fu Lii, “An
Efficient Folded Architecture for Lifting-Based Discrete
Wavelet Transform”, IEEE Transactions on Circuits and
Systems-II: Express Briefs, 200956 (4), pp. 290-294.
[7] Chengyi Xiong, Jinwen Tian, and Jian Liu, “Efficient
Architectures for Two-Dimensional Discrete Wavelet
Transform Using Lifting Scheme” ,IEEE Transactions on
Image Processing, 2007, 16 (3), pp. 607-614.
[8] Chih-Hsien Hsia, Jen-Shiun Chiang, Member, and Jing-
Ming Guo “Memory-Efficient Hardware Architecture of 2-D
Dual-Mode Lifting-Based Discrete Wavelet Transform”,
IEEE Transactions on Circuits and Systems for Video
Technology, 2013, 23 (4), pp. 671-683.

IRJET- An Efficient VLSI Architecture for 3D-DWT using Lifting Scheme

More Related Content

What's hot (19)

Similar to IRJET- An Efficient VLSI Architecture for 3D-DWT using Lifting Scheme (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET- An Efficient VLSI Architecture for 3D-DWT using Lifting Scheme