Automatic Image Registration Using 2D-DWT

International Journal of Engineering Science Invention
ISSN (Online): 2319 – 6734, ISSN (Print): 2319 – 6726
www.ijesi.org ||Volume 4 Issue 2 ||February 2015 || PP.50-52
www.ijesi.org 50 | Page
Automatic Image Registration Using 2D-DWT
1,
M.Deepa, 2,
Dr.T.V.U.Kirankumar
Bharath University, Chennai, India.
ABSTRACT : Image registration is an important and fundamental task in image processing used to match
two different images. Image registration estimates the parameters of the geometrical transformation model that
maps the sensed images back to its reference image. A Feature-Based Approach to automated image-to-image
registration is presented. In this paper, various methods are used in different Phases of Image registration. The
characteristics of this approach is it combines scale interaction of Discrete wavelets for feature extraction,
Scale Invariant Feature Transform (SIFT) for feature matching. Scale-invariant feature transform (or SIFT) is
an algorithm in computer vision to detect and describe local features in images. SIFT feature descriptor is
invariant to uniform scaling, orientation, and partially invariant to affine distortion and illumination changes.
KEY WORDS: Image Registration, Discrete Wavelet Transform, Scale Invariant Feature Transform
I. INTRODUCTION
In this paper, we propose and implement a novel image registration method to register image with large
rotation, scaling and translation. In all types of image registration, Robustness of the algorithm is the main and
required goal. However, due to diversification of images acquired their contents and purpose of their alignment,
it is almost impossible to design universal method for image registration that fulfill all requirements and suits all
types of applications. Many of the image registration techniques have been proposed and reviewed. Image
registration techniques can be generally classified in two categories. Intensity based and feature based. The first
category utilizes image intensity to estimate the parameters of a transformation between two images using an
approach involving all pixels of the image. In second category a set of feature points extracted from an image
and utilizes only these extracted feature points instead of all whole image pixels to obtain the transformation
parameters. In this paper, a new algorithm for feature based image registration is proposed. The proposed
algorithm is based on three main steps, feature point extraction, correspondence between extracted feature points
and transformation parameter estimation.
II. IMAGE REGISTRATION BLOCK DIAGRAM
Image registration typically consists of the following steps:
Fig 1 Block Diagram of Image Registration

Automatic Image Registration using…
III. DISCRETE WAVELET TRANSFORM
The Discrete Wavelet Transform (DWT), which is based on sub-band coding, is found to yield a fast
computation of Wavelet Transform. It is easy to implement and reduces the computation time and resources
required. The two-dimensional DWT of an image function s(n1, n2) of size N1 x N2 may be expressed as
      





1
0
21,,
1
0
21
21
210
1
1
210
2
2
,,
1
,,
N
n
kkj
N
n
nnnns
NN
kkjW 
      





1
0
21
1
0
21
21
210
1
1
2,1,0
2
2
,,
1
,,
N
n
i
N
n
i
nnnns
NN
kkjW kkj

Where i = {H, V, D} indicate the direction index of the wavelet function. As in one-dimensional case j0
represents any starting scale, which may be treated as j0=0. Given the above two equations are two-dimensional
DWT.
The flow chart of the Discrete wavelet transform sub band coding on the digital image is shown in
figure 2, here L refers to low frequency component, H refers to high frequency and the number 1 and 2 refer to
the decomposition level of the Discrete wavelet transform. The result of the 2-D Discrete Wavelet Transform
from level one to level three is shown in figure 3. The sub image LL is the low frequency component, it is the
approximate sub image of the original image; the sub image HL is the component of the low frequency in
horizontal direction and the high frequency in vertical direction, it manifests the horizontal edge of the original
image; the sub image LH is the component of the high frequency in horizontal direction and the low frequency in
vertical direction, it manifests the vertical edge of the original image; the sub image HH is the high frequency
component, it manifests the oblique edge of the original image. It is shown that most energy of the original
image is contained in the LL2 low frequency region. And the other region in the same size reflect edge feature of
the image in different angles. Here we use the 2-D Discrete Haar wavelet transform for the decomposition of the
images.
Fig. 2 Flow chart of the DWT sub band coding
IV. FEATURE EXTRACTION
Our proposed algorithm is based on feature based image registration. So first and important task we
have to do is feature extraction using Discrete Wavelet Transforms. For the image registration point of view, the
extracted feature points must have to fulfill some basic objectives like they should be robust enough to some
level of variations in scale and the feature extractor should have the ability to be modified to adapt image
structures at different scales and the extracted features should have a well-localized support in the image.
Approximated Features from both images are extracted using Discrete Wavelet Transform and Control point
correspondence and Feature matching is achieved using Scale-Invariant Feature Transform (SIFT). The scale
invariant feature transform (SIFT) algorithm for image features generation which are invariant to image
translation, scaling, rotation and partially invariant to illumination changes and affine projection. After extracting
the frequency component from the Discrete Wavelet Transform, following steps are applied to detect the features
and match the corresponding features from sensed and referenced image. The next process is selecting the
control points using the control selection tool that estimates the match of the control points based on the
geometric relationship of the previously selected control points.
V. SIMILARITY MATCHING
All the previous steps such as extracting approximated level coefficients, locating and extracting the
Interest Points are doing for both reference and sensed Images. Now, we have features of Image 1 and features
of Image 2. Next step is to match the features of both images. By using the match Features function, we will get

Automatic Image Registration using…
Index Pairs, containing P pairs of indices. These indices contain features most likely to correspond between the
two input feature sets. Next step is to retrieve the locations of corresponding points for each image from the
obtained Index Pairs. Our proposed Algorithm uses Nearest Neighbour Ratio method for finding true and false
matches. Then the images will be registered automatically.
VI. EXPERIMENT RESULT
The experiment result are shown below fig 2(a) shows the reference image fig 2(b) shows the feature
control points fig 2(c) shows the sensed image fig 2(d) shows the control point for sensed images these control
points are matched using shift invariance feature transform if it matches the images are registered.
Fig.2. (a) Reference Image, (b) Reference Image Feature points, (c) Sensed Image, (b) Sensed Image Feature
points
VII. CONCLUSION
In this paper, a new feature based image registration approach is proposed. This type of approach gives
a new dimension to the existing feature based image registration methods. After successful feature extraction
using Discrete Wavelet, correspondence between extracted feature points is established using Scale Invariant
Method. With this proposed algorithm, we registered images with any degree of rotation and scaling.
REFERENCES
[1]. Gang Hong and Yun Zhang, “Combination of feature-based and area based image registration technique for high resolution remote
sensing image,” Geoscience and Remote Sensing Symposium, 2007, IGARSS 2007, pp.377-380, 2007
[2]. Hongli Shi and Shuqian Luo, "Image registration using the shift-insensitive discrete wavelet transform", in Proceedings of the IEEE
International Conference on Medical Image Analysis and Clinical Applications (MIACA), 2010
[3]. Jignesh N Sarvaiya and Dr. Suprava Patnaik, ”Automatic Image Registration Using Mexican Hat Wavelet, Invariant Moment, and
Radon Transform”, (IJACSA) International Journal of Advanced Computer Science and Applications, Special Issue on Image
Processing and Analysis, 2011
[4]. Rajaei, A, "Feature Extraction of Currency Notes: An Approach Based on Wavelet Transform" in Proceedings of the IEEE
International Conference on Advanced Computing & Communication Technologies (ACCT),2012
[5]. C.Wang,R.X.Gao, Wavelet transform with spectral post-processing for enhanced feature extraction, IEEE Instrumentation and
Measurement Technology Conference, 2003.
[6]. A.Quddus and O. Basir, “Wavelet-Based Medical Image Registration for Retrieval Applications,” in Proceedings of the 2008
International Conference on Bio-Medical Engineering and Informatics-Volume 02, pp. 301–305, IEEE Computer Society, 2008.
[7]. M. S. Yasein and P. Agathoklis, "A robust, feature-based algorithm for aerial image registration," in Proceedings of the IEEE
International Symposium on Industrial Electronics (ISIE 2007), Vigo, Spain, pp.1731-1736, 2007
[8]. B.Zitova and J. Flusser, “Image registration methods: a survey,” Image and vision computing 21(11), pp. 977–1000, 2003.
.

Automatic Image Registration Using 2D-DWT

More Related Content

What's hot (19)

Similar to Automatic Image Registration Using 2D-DWT (20)

Recently uploaded (20)

Automatic Image Registration Using 2D-DWT