Docstoc

NOVEL IMAGE FUSION TECHNIQUES USING GLOBAL AND LOCAL KEKRE WAVELET TRANSFORMS

Document Sample
NOVEL IMAGE FUSION TECHNIQUES USING GLOBAL AND LOCAL KEKRE WAVELET TRANSFORMS Powered By Docstoc
					 INTERNATIONAL JOURNAL OF COMPUTER ENGINEERING
  International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-
                             & TECHNOLOGY January- February (2013), © IAEME
  6367(Print), ISSN 0976 – 6375(Online) Volume 4, Issue 1,(IJCET)

ISSN 0976 – 6367(Print)
ISSN 0976 – 6375(Online)
Volume 4, Issue 1, January- February (2013), pp. 89-96
                                                                      IJCET
© IAEME: www.iaeme.com/ijcet.asp
Journal Impact Factor (2012): 3.9580 (Calculated by GISI)        ©IAEME
www.jifactor.com




    NOVEL IMAGE FUSION TECHNIQUES USING GLOBAL AND
           LOCAL KEKRE WAVELET TRANSFORMS


                                Dr. Sudeep D. Thepade
                    Professor, Department of Computer Engineering,
                    Pimpri Chinchwad College of Engineering, Pune

                                Mrs. Jyoti S.Kulkarni
                   Senior Lecturer,Department of Information Tech.,
                   Pimpri Chinchwad College of Engineering, Pune



  ABSTRACT

         Image Fusion is the process of combining the information from multiple
  images such that the fused image gives or represents more information than that of
  single image gives. Images for image fusion may be from single sensor with different
  time slots or may be from multiple sensors. Image fusion is used in different
  applications such as medical imaging, Military images, Multisensory images,
  Multifocus images etc. Different transforms are used for image fusion. In this
  proposed method, Kekre transform is used. Kekre transform is one of the orthogonal
  transforms. Here Kekre transform along with Local Kekre Wavelet Transform and
  Global Kekre Wavelet Transform are proposed to be used in novel image fusion
  methods. For each of the proposed Image Fusion techniques , the Average , Minimum
  and Maximum are considered for generation of fused image. Experimentation is
  performed on ten sets of images to generate the fused images. Result has shown the
  Local Kekre wavelet transform proves to be better for image fusion than Global Kekre
  wavelet transform and Kekre Transform. Also the averaging based fusion is better
  than minimum or maximum.

  Keywords- Kekre Transform, Local Transform, Global Transform


                                           89
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-
6367(Print), ISSN 0976 – 6375(Online) Volume 4, Issue 1, January- February (2013), © IAEME

I. INTRODUCTION

        Now a days, Image fusion is used in many fields. In this, the images from different
sensors or from same sensor with different time or season are fused together. After image
fusion, the resultant image generated will be more informative than source images. In fused
image, the relevant information from source images get used to find more information which
will be further processed.
        There are increasing applications of image fusion in different fields. In topographic
mapping, the area that is not covered by one sensor might be available in another sensor. By
combining the information from these sensors, the area or map is updated. Similarly to get
the idea about natural hazards such as flood monitoring and snow monitoring. Image fusion is
also used in geology to get the information on soil geochemistry, vegetation, land use, soil
moisture and surface roughness.

       Many image fusion methods are available such as Intensity Hue Saturation (IHS),
Principle Component Analysis (PCA), Brovey Transform (BT), High Pass Filtering (HPF) ,
High Pass Modulation (HPM) and Transform domain image fusion..
       Here use of Kekre Transform is proposed along with Local Kekre Wavelet Transform
and Global Kekre Wavelet Transform for Image Fusion.

       Section II describes the Kekre Transform and generation of Local Kekre Wavelet
Transform and Global Kekre Wavelet Transform. Section III describes the proposed Image
Fusion Method with block diagram. Section IV describes the experimentation on proposed
Image Fusion Method and Section V describes the results and discussion on the fused images
by using these methods.

II. USED TRANSFORMS

(a) Kekre Transform: In Kekre Transform, it is not essential that the matrices have to be in
powers of 2. The Kekre Transform is generated by using equation (1).
                    1            , x≤y
              {
   Kx,y =           (-N+(x+1) , x=y+1
                    0            , x>y+1
                                                                                          (1)
       Let the following matrix generated by this equation having PxP size.

                                         1   2    …    P
                                    1             …
                                    2             …
                                                  …
                                    P             …

                            Figure 1: PxP Kekre Transform matrix




                                                 90
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-
6367(Print), ISSN 0976 – 6375(Online) Volume 4, Issue 1, January- February (2013), © IAEME

(b) Local Kekre WaveletTransform : Generation of Local Kekre Wavelet Transform of size
P2xP2 from Kekre Transform of size PxP is shown in fig. 2.

                              …               …        …            …
                              …               …        …            …
                              …               …        …            …
                              …               …        …            …
                              …       0   0   …    0   …    0   0   …   0
                  0       0   …   0           …        …    0   0   …   0

                  0       0   …   0   0   0   …    0   …            …
                              …               …        …            …
                              …       0   0   …    0   …    0   0   …   0
                  0       0   …   0           …        …    0   0   …   0

                  0       0   …   0   0   0   …    0   …            …

                 Figure 2: P2xP2 matrix of Kekre Local Wavelet Transform

(c) Global Kekre Wavelet Transform: Generation of Global Kekre Wavelet Transform of size
P2xP2 from Kekre Transform of size PxP is shown in fig. 3.

                              …               …        …            …
                              …               …        …            …
                              …               …        …            …
                              …               …        …            …
                              …       0   0   0    0   0    0   0   0   0
                              …       0 0 0        0 0      0 0 0       0
                              …       0 0 0        0 0      0 0 0       0
                      0   0   0   0                  0      0 0 0       0
                      0 0 0       0           …         0   0 0 0       0
                      0 0 0       0           …        0    0 0 0       0
                      0 0 0       0 0 0 0          0        0   0   0   0
                      0 0 0       0 0 0 0          0 0              …
                      0 0 0       0 0 0 0          0 0              …
                      0 0 0       0 0 0 0          0 0              …

                Figure 3: P2xP2 matrix of Kekre Global Wavelet Transform

The normalization of these wavelet transforms done before the use in processing.



                                              91
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-
6367(Print), ISSN 0976 – 6375(Online) Volume 4, Issue 1, January- February (2013), © IAEME

III. PROPOSED IMAGE FUSION METHOD
       The proposed image fusion method using Kekre wavelet transform is given in fig.4.

                 Apply Local /
                                     Transformed
   Image 1       Global Kekre                           Fusion using
                                     Image
                 Transform                              Minimum/M          Inverse
                                                                                        Fused
                                                        aximum/Ave         Transform
                                                                                        Image
                                                        rage
                Apply Local /                           Method
   Image 2      Global Kekre        Transformed
                Transform           Image



                Figure 4: Basic Block Diagram of proposed Image Fusion Method
In the proposed Image Fusion Method, normalized local or global transform is applied to two
blurred images separately. Three coefficients are available after this transformation. The
inverse transform is applied to find fused image. The coefficients are compared to find better
fused image.
IV. IMPLEMENTATION / EXPERIMENTATION
       For experimentation, set of six images are considered to fused using proposed image
fusion method as shown in fig 5.




              Set1: Tulip                                 Set2: Moon




              Set3: Animal                                   Set4: Apple




                Set5: Wool                                 Set6: Scene

                         Figure 5: Test bed of set of images to be fused

                                              92
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-
6367(Print), ISSN 0976 – 6375(Online) Volume 4, Issue 1, January- February (2013), © IAEME

V. RESULTS AND DISCUSSION

        The Kekre Transform, Local Kekre Wavelet Transform and Global Kekre Wavelet
Transform is applied on the images given in fig.5. Then the fused image is compared with
original image to find the mean square error. Result generated shows that local transform is
better than orthogonal transform as well as global transform.

Table1. Comparison of Kekre Transform, Kekre Local Wavelet transform and Kekre Global
Transform


                                   Kekre Wavelet
                                                       Kekre Wavelet                    Kekre
                                       Local
                                                      Global Transform                transform
                                    Transform


               Tulip                  271.7474              275.9028                    17176

               Moon                    91.3706              91.1495                     24568

              Animal                  122.5144              122.5086                    13581

               Apple                  363.9339              363.9184                    18797

               Wool                   385.4935              397.9610                    15315

               Scene                  191.2848              190.5862                    15438

             Average                   237.72                240.34                   17479.17


                                           Comparison of KT, LKWT,GKWT
                                   30000

                                   25000
                                                                      Kekre transform
                       MSE VAlue




                                   20000

                                   15000
                                                                      Kekre Wavelet
                                   10000                              Global
                                                                      Transform
                                    5000
                                                                      Kekre Wavelet
                                       0                              Local
                                                                      Transform




            Figure 6: Graphical representation of comparison between transforms.



                                                       93
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-
6367(Print), ISSN 0976 – 6375(Online) Volume 4, Issue 1, January- February (2013), © IAEME

        By considering the Kekre Local wavelet Transform, the minimum, maximum and
average values are compared. After comparison it shows that average value is better than
other two values.

Table2. Comparison of Average, Minimum and Maximum values in Local KekreWavelet
transform



                                             Average         Minimum                Maximum



                        Tulip                271.7474            326.2178           327.652

                      Moon                   91.3706             102.0376           105.223

                   Animal                    122.5144            143.7189           148.514

                     Apple                   363.9339            413.4674           440.807

                       Wool                  385.4935            474.4746           479.406

                      Scene                  191.2848            217.3164           230.101



                               500


                               400
                   MSE value




                               300
                                                                                       Average
                               200
                                                                                       Minimum
                                                                                       Maximum
                               100


                                0
                                     Tulip    Moon Animal Apple     Wool    Scene

                                                        Images




 Figure 7: Graphical representation of comparison between values of Kekre Local Wavelet
                                        transforms.




                                                           94
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-
6367(Print), ISSN 0976 – 6375(Online) Volume 4, Issue 1, January- February (2013), © IAEME


The output images after the application of these three transforms is given in figure 8.




                        (a)Original       (b) Blurred     (c) Blurred
                            Image              Image          Image




                        (d) Average     (e)           (f) Maximum
                                       Minimum
                               Kekre Local Wavelet Transform




                        (g) Average     (h) Minimum (i) Maximum
                                       Kekre Transform




                        (j) Average    (k) Minimum (l) Maximum
                              Kekre Global Wavelet Transform

                  Figure 8: Input and Output Images of all the Transforms.


Figure 8 represents the input and output images of Kekre Transform, Kekre Local wavelet
transform and Kekre Global Wavelet Transform. In figure, first row are the original image
and blurred images respectively. Second row represents outputs of Kekre Local Wavelet
Transform with Minimum, Maximum and Average values respectively. Third row represents
outputs of Kekre Transform with Minimum, Maximum and Average values respectively. And
last row represents, outputs of Kekre Global Wavelet Transform with Minimum, Maximum
and Average values respectively. From this, the Kekre Local Wavelet transform gives better
result than remaining two methods.

                                               95
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-
6367(Print), ISSN 0976 – 6375(Online) Volume 4, Issue 1, January- February (2013), © IAEME

CONCLUSIONS

       Image Transform based fusion is gaining momentum in imaging research. Here novel
Image Fusion methods are proposed using Kekre Transform, Local Kekre Wavelet Transform
and Global Kekre Wavelet Transform .Experimentation has shown that the Kekre Wavelet
Transform based fusion is outperforming. In all Local Kekre Wavelet Transform with
averaging gives better performance for image fusion.

REFERENCES

[1]H.B.Kekre, Tanuja K. Sarode, Sudeep D.Thepade, Sonal Shroff,”Instigation of Orthogonal
Wavelet Transforms using Walsh, Cosine, Hartley, Kekre Transforms and their use in Image
Compression”,International Journal of Computer Science and Information Security, Vol.9,
No.6,2011,125-133.
[2]H.B.Kekre, Archana Athawale, Dipali Sadavarti,”Algorithm to generate Kekre’s Wavelet
Transform from Kekre’s Transform”,International Journal of Engineering Science and
Technology, Vol.2(5),2010,756-767.
[3] H.B.Kekre, Tanuja K. Sarode, Sudeep D.Thepade, ,”Inception of Hybrid Wavelet
Transform using two Orthogonal Transforms and it’s use for Image
Compression”,International Journal of Computer Science and Information Security, Vol.9,
No.6,2011,80-87.
[4] Peijun Du,Sicong Liu, Junshi Xia, Yindi Zhao,”Information Fusion Techniques for
change detection from multi temporal remote sensing images”
[5]Yufeng Zheng,Edward A Essock,”Alocal coloring method for night vision colorization
utilizing image analysis and fusion”, Information Fusion 9(2008) 186-199.
[6]Chandan Singh, Pooja,”An effective image retrieval using the fusion of global and local
transforms based features”, Optics and Laser Technology 44(2012), 2249-2259.
[7]Yong Yang, “A Novel DWT based multi focus image fusion method”, Procedia
engineering 24(2011) 177- 181.
[8]Youdong Ding, Cai Xi, Xiaocheng Wei, Jianfei Zhang,”A new framework for image
completion based on image fusion technology”, Procedia engineering 29(2012) 3826-3830.
[9]Tao Wu, Xiao-Jun Wu, Xiao-Qing Luo,”A study on fusion of different resolution images”,
Procedia engineering 29(2012) 3980-3985.
[10] Deng Minghui, Zeng Qingshung, Zhang Lanying,”Research on Fusion of Infrared and
visible images based on directionlet transform”, IERI procedia 3(2012) 67-72.
[11] B.V. Santhosh Krishna, AL.Vallikannu, Punithavathy Mohan and E.S.Karthik Kumar,
“Satellite Image Classification Using Wavelet Transform” International journal of
Electronics and Communication Engineering &Technology (IJECET), Volume 1, Issue 1,
2010, pp. 117 - 124, Published by IAEME.
[12] B.K.N.Srinivasa Rao and P.Sowmya, “Architectural Implementation Of Video
Compression Through Wavelet Transform Coding And Ezw Coding” International journal of
Electronics and Communication Engineering &Technology (IJECET), Volume 3, Issue 3,
2012, pp. 202 - 210, Published by IAEME.
[13] Hitashi and Sugandha Sharma, “Fractal Image Compression Scheme Using
Biogeography Based Optimization On Color Images”, International journal of Computer
Engineering & Technology (IJCET), Volume 3, Issue 2, 2012, pp. 35 - 46, Published by
IAEME.


                                           96

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:4
posted:2/2/2013
language:
pages:8