Sei sulla pagina 1di 22

Mini Project Seminar

on
Implementation of Fusing Images With Different Focuses using DWFT

Presented By
Komalakumari
USN:4JN12LDS05

Under the guidance of


Dr.Manjunatha.P

Department of Electronics and Communication Engineering


J.N.N. College of Engineering, Shimoga

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 1 / 22
Contents Contents

Contents

1 Introduction
2 Aim of the project
3 Implementation Steps
4 Implementation details
5 Results
6 Conclusion
7 References

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 2 / 22
Introduction

Image Fusion
Image Fusion is a process of combining the relevant information from
a set of images, into a single image, where the resultant fused image
will be more informative and complete than any of the input images[?]
Image fusion can be divided into three categories
1 Pixel level
It generates a fused image in which information associated with each
pixel is determined from a set of pixels in source images
2 Feature level
Feature-based fusion at feature level requires an extraction of objects
recognized in the various data sources
3 Decision level
Decision-level fusion consists of merging information at a higher level
of abstraction, combines the results from multiple algorithms to yield
a final fused decision

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 3 / 22
Introduction contd..

All these fusion levels involve loss of information in the information


extraction process, which consequently leads to less accurate fusion
results and reduced contrast image
Hence multiscale transforms is used-DWT, DWFT,Laplacian
pyramid,etc

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 4 / 22
Introduction contd..

Figure: Reference paper

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 5 / 22
Aim of the project

Aim of the project

The objective of this project is to implement the fusing images with


different focuses of the same scene in order to obtain an
everywhere-in-focus image by the use of discrete wavelet frame transform(
DWFT).

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 6 / 22
DWFT

DWFT
DWF scheme splits the input sequence into the wavelet frame
sequence wi (k) which is stored.
the scale frame sequence si (k) which serves as input for the next
decomposition level[1].
wi+1 (k) = [g ]↑2i ∗ si (k)
si+1 (k) = [h]↑2i ∗ si (k)
The reconstruction formula is
∼ P ∼
f = h ∗si (k) + gi ∗ wi (k)
i

Figure: Discrete wavelet frame decomposition of the ith level

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 7 / 22
DWFT

The implementation of discrete wavelet frame transform(DWFT) is


similar to that of DWT in which there is no downsampling operation.
In the decomposition phase of 2-D DWFT, each row of the image is
separately filtered by h and g.
The resulting row-transformed image is then similarly filtered in the
column direction, finally yielding four subbands at the first
decomposition level (i=1).
The three detail subbands, D i LH, D i HL, D i HH, contain the vertical,
horizontal, and diagonal high frequency information, respectively,
while the approximation subband S i LL is a low pass filtered version of
the original image[2].
This approximation subband is subsequently passed to the next level
for further subband decomposition.
Thus, a DWFT with N decomposition levels will have a total of
3N + 1 frequency subbands, all of them are of the same size.

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 8 / 22
Block diagram

Figure: Block Diagram

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 9 / 22
Implementation Steps

Implementation Steps
1 Step 1:
Several source images with different focuses of the same
scene are taken
2 Step 2:
Decompose the two source images by applying DWFT,
resulting in a total of details subbands and one approximation
subband.
3 Step 3:
Perform average of approximation coefficients of both
decomposed images.
4 Step 4:
Apply maximum selection scheme(Choose-max Rule) to the
detailed coefficients.
5 Step 5:
Finally, the fused image is obtained by performing the
inverse DWFT on the fused coefficients[3].
Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 10 / 22
Implementation Steps

Fusion Rules
The fusion rule includes two steps
Activity level measurement:
The activity level of wavelet coefficients reflects the local energy
in the space.
There are three categories of methods for computing the activity
level AI (p) at position p.
1 coefficient-based
2 window-based
3 region-based measures
Coefficient-based activity (CBA):
In CBA, the activity level is given by
AI (p) = |DI (p)| or AI (p) = |DI (p)|2
The index of each DWT coefficient is denoted by p = (m, n, k, l).
DI (p) and AI (p) are the value and activity level of p DWT coefficient.
Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 11 / 22
Implementation Steps

Coefficient combining method:


Selection and averaging are probably the most popular coefficient
combining methods[4].
Selection method can collect the largest DWT coefficients between
two images.
Therefore, it is suitable for collecting the edges or corners, i.e.
detailed information.
The simplest selection method is choose-max (CM)

DX (p), if AX (p) ≥ AY (p)
Dz (p) =
DY (p), if AX (p) ≤ AY (p)

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 12 / 22
Implementation Steps

Averaging method is employed for combing the low-frequency bands


because the approximation images of the images to be fused usually
look similar.
The value of the pixel p(i, j) of each image is taken and added. This
sum is then divided by 2 to obtain the average. The average value is
assigned to the corresponding pixel of the output image which is
given by
K (i, j) = {X (i, j) + Y (i, j)}/2
where X (i, j) and Y (i, j) are the two input images, and this is
repeated for all pixel values[5].

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 13 / 22
Implementation Steps

Figure: implementation of Image fusion

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 14 / 22
Implementation Details
1 Performing DWFT for images taken and features extraction using
wavelet coefficients
2 Perform inverse DWFT
All the above steps are implemented in matlab

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 15 / 22
Results

Results

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 16 / 22
Results

Figure: Left focused image and its DWFT

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 17 / 22
Results

Figure: Right focused image and its DWFT

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 18 / 22
Results

Figure: The Fused image


Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 19 / 22
Conclusion

Conclusion

The fusing of the images with different focuses using Discrete wavelet
frame transform(DWFT) is implemented, which retains the important
features from both the images.

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 20 / 22
References

References
J. T. Kwok and Y. Wang, “Using the discrete wavelet frame transform
to merge landsat tm and spot panchromatic images,”
www.elsevier.com/locate/inffus, 2002.
H. Wang, Zhongliang, and J. J. Li, “An image fusion approach based
on discrete wavelet frame,” Institute of Aerospace Information and
Control, Shanghai Jiao Tong University, Shanghai, P. R. China, 2003.
S. Li and Y. Wang, “Fusing images with different focuses using
support vector machines,” IEEE TRANSACTIONS ON NEURAL
NETWORKS, vol. 15, no. 6, November 2004.
S.-G. Huang, “Wavelet for image fusion,” Graduate Institute of
Communication Engineering and Department of Electrical
Engineering, National Taiwan University.
D. K. Sahu and M.P.Parsai, “Different image fusion techniques a
critical review,” International Journal of Modern Engineering Research
(IJMER),
Komalakumari vol. 2,
(2nd sem, M.Tech 2012.
DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 21 / 22
Thank You

Komalakumari (2nd sem, M.Tech DECS) Fusing Images With Different Focuses using DWFT June 21, 2013 22 / 22

Potrebbero piacerti anche