Scholars Research Library
Archives of Applied Science Research, 2013, 5 (2):203-207 (http://scholarsresearchlibrary.com/archive.html)
ISSN 0975-508X CODEN (USA) AASRC9
Generalization of two dimensional canonical SS-transform
S. B. Chavhan
Department of Mathematics & Statistics, Yeshwant Mahavidyalaya, Nanded (India)
_____________________________________________________________________________________________
ABSTRACT
The canonical transform (CT) is a widely used integral transform with many applications in applied mathematics, mathematical physics and engineering sciences. The canonical transform (CT) is related to the Fourier transform (FT). The Fourier transform can be generated into the fractional Fourier transform, linear canonical transform.
This paper is devoted for analytic study of two dimensional generalized canonical SS-transform, and parsevals identity for two dimensional canonical SS-transform.
AMS Subject code: 46F12 and 44
Keywords: 2D canonical cosine-cosine transform, 2D-canonical sine-sine transform, generalized functions.
_____________________________________________________________________________________________
INTRODUCTION
The Fourier transform (FT) is undoubtedly one of the most valuable and frequently used tools in signal processing and analysis [2]. The fractional Fourier transform has been found to have several applications in the areas of optics and filtering etc [4], [5]. Fractional Fourier transforms special case of linear canonical transform, numbers of integral transform are extended in its fractional domain for example, Almeida [1] had studied fractional Fourier transform, Chavhan.S.B and Borkar.V.C [3] developed two dimensional canonical cosine-cosine transform. The linear canonical transform is defined as [6]
2 ( / )2
2 2
{ ( )}( ) 1 ( ) ,
2
i d s i s t i a b t
b b
C Tf t s e e e f t dt
πib
∞ −
−∞
= ⋅
∫
⋅This paper emphasizes define two dimensional canonical SS-transform, and deriving its analyticity theorem then some properties of two dimensional of SS-transform are discussed and finally conclusion is given. Notation and terminology as per zemanian [7],[8].
2 Definition two dimensional (2D) canonical SS- transform:
{
2DCSST f t x( )
,} ( )
s w, = f t x K( )
, , s1( ) ( )
t x K, s2 x w,{ ( ) } ( )
( )
2 2 2 2 2 2 2( )
2 , ,
1 1
1 sin sin ,
2 2
i d i d i a i a
s w t x
b b b b
DCST f t x s w
s w
e e t x e e f t x dxdt
b b
ib ib
π π
2 ∞ ∞
−∞ −∞
= − ∫ ∫
Where, 1
( ) ( )
2 22 2
, 1 sin .
2
i d i a
s t
b b
s
K t s i e st e
ib b π
= −
when b≠0 ( )2
( )
2 ic ds
d e δ t ds
= − when b = 0
( ) ( )
2 22
2 2
, 1 sin .
2
i d i a
w x
b b
s
K x w i e wx e
ib b π
= −
when b≠0 ( )2
( )
2 ic dw
d e δ x dw
= − when b = 0
where E k,
{
Ks1( ) ( )
t s K, s2 x w,}
sup t D D Ktk xl s1( ) ( )
t s K, s2 x w,x
γ = −∞ < < ∞ < ∞
−∞ < < ∞
3 Analyticity Theorem of 2D Canonical SS-Transform
Theorem 3.1 :( Analyticity) Let f∈E R1( n) and its two canonical SS- transform be defined by,
{ }
2 2 2 2
2 2 2 2
2 ( , ) ( , )
1 1
sin sin ( , )
2 2
i d i d i a i a
s w t x
b b b b
DCSST f t x s w
s w
e e e e t x f t x dxdt
ib ib b b
π π
∞
−∞
= −
∫
Then
{
2DCSS T f t x( , ) ( , )}
s w is analytic on Cn,if the a, b, sup pf ⊂saand sbwhere{
: n, , 0}
sa = t t∈R t ≤a a> ,sb=
{
x x: ∈Rn, x ≤b b, >0}
moreover{
2DCSST f t x( , ) ( , )}
s w is differentiable and D Dsk lw{2DCSST f t x( , ) ( , )} s w = f t x D D K( , ), sk wl s1( , )t s Ks2( , )x w Proof: Let, s: {s1, s2,………sj……sn}∈
Cn andw: {w1, w2,………wj……wn}
∈
CnWe first prove that, {2 ( , )}( , ) exists,
j j
DCSS T f t x s w
s w
∂ ∂
∂ ∂
1 2
{2 ( , )}( , ) ( , ), ( , ) ( , ) (3.1)
n n n n
s s
n n n n
j j j j
DCSST f t x s w f t x K t s K x w
s w s w
∂ ∂ = ∂ ∂
∂ ∂ ∂ ∂
we prove the result n = 1, the general result following by induction.
For fixed sj ≠ 0 choose two concentric circles C and C1 with centre sj and radii r and r1 respectively such that 0<r<r1< |sj|.
Let
∆ s
j be a complex increment satisfying0 < ∆ s
j< r
. Also for fixed wj ≠ 0.Again choose two concentric circles C and C1 with centre wj and radiir' and r1' respectively such that 0<r' <r1' < |wj|.Let ∆wj be a complex increment satisfying 0< ∆wj <r'
______________________________________________________________________________
Consider,
1 2
(2 ) ( , ) (2 ) ( , ) (2 ) ( , ) (2 ) ( , )
( , ), ( , ) ( , )
( , ), ( ) ( ) (3.2)
j j j j j j j j j j
j j
s s
j j
j j
DCSST s s w DCSST s w DCSST s w w DCSST s w
s w
f t x K t s K x w
s w
f t x s t w x
+ ∆ − + ∆ −
∆ ∆
− ∂ ∂
∂ ∂
= Ψ∆ Ψ∆
where
1 1 2 1
( ) ( ) 1 ( , , ) ( , )
j j s j j n s
j
s t w x K t s s s s s K t s
s
Ψ∆ ∆ =∆ LL + ∆ L −
2 1 2 2
1 s ( , , j j n) s ( , )
j
K x w w w w w K x w
w
+ ∆ −
∆ LL L
1( , ) 2( , )
n n
s s
n n
j j
K t s K x w
s w
− ∂ ∂
∂ ∂
For any fixed
( )
t x, ∈Rn and any fixed integer.1 2 0
( , n)
k= k k LLk ∈N and l=( ,l l1 2LLln)∈N0
1( , ) 2( , )
k l
t x s s
D D K t s K x w is analytic inside and onC' and C'1. By Cauchy integral formula.
( ) ( )
1 2
' ' 1
2 2 2
2
( , )
1 1 1 1 1
, ,
4 ( )
1 1 1 1
( )
k l
t x j j
k l
t x s s
j j j j j
c c
j j j j j
D D s w t x
D D K t s K x w
s z s s z s
i z s
w y w w y w y w
π Ψ∆ ∆
= ∆ − − ∆ − − − −
− −
∆ − − ∆ − −
∫ ∫
dzdy
s = (s1………sj-1, z, sj+1………sn).
w = (w1………wj-1, y,wj+1………wn).
1 2
' ' 1
2 2 2
( , ) ( , )
4 ( ) ( ) ( ) ( ) ,
k l
t x s s
j j
j j j j j j
c c
D D K t s K x w s w
z s s z s y w w y w dzdy
π
=∆ ∆
−
∫ ∫
− − ∆ − − − ∆ −But for all z∈C' and y∈C1' and
( )
t x, restricted to a compact subset of Rn,1( , ) 2( , )
k l
t x s s
D D K t s K x w is bounded by constant Q.
' ' 1
2 ' '
1 1 1 1
( , )
4 ( ) ( )( ) ( )
j j
k l
t x j j
c c
s w Q
D D s w t x dz dy
r r r r r r
π Ψ∆ ∆ ≤ ∆ ∆
− −
∫ ∫
2 ' '
1 1 1 1
4 ( ) ( )( ) ( )
j j
s w Q
r r r r r r
π
∆ ∆
≤ − −
Thus as ∆ →sj 0,and ∆wj →0, D Dtk lxΨ∆ ∆sj w t xj( , )tends to zero uniformly on the compact subset of Rn.,therefore it follows that Ψ∆ ∆sj w t xj( , ) converges in E(Rn) to zero. Since f∈E1, we conclude (3.2) tends to zero. Therefore {2DCSST f t x( , ) ( , )} s w is differentiable with respective sj and wj. But this is true for all i, j=1,2,……n .Hence
{ }
and, D Dsk wl{2DCSST f t x( , ) ( , )} s w = f t x( , ),D D Ksk wl s1( , )t s Ks2( , )x w . 4 Parsevals Identities for Two dimensional Canonical SS-Transform.
If f t x
( )
, and g t x( )
, are inversion of two dimensional Canonical SS-Transform{2DCSST}( , )s wand {2DCSST}( , )s w respectively, then
4.1 ∞ ∞ f t x g t x dxdt( ) ( ), , =4π2∞ ∞{2DCSST}( ){s w, 2DCSST}( )s w dsdw,
−∞ −∞
∫ ∫
−∞ −∞∫ ∫
4.2 ∞ ∞ f t x( ), 2dxdt=4π2∞ ∞ {2DCSST}( )s w, 2dsdw
−∞ −∞
∫ ∫
−∞ −∞∫ ∫
Proof: By definition two dimensional Canonical SS-Transform,
{ ( )}( )
( ) 2 2 2 2 2 2 2 ( )
2 , ,
1 1
1 sin sin , ....(4.1)
2 2
i d i d i a i a
s w t x
b b b b
DCST g t x s w
s w
e e t x e e g t x dxdt
b b
ib ib
π π
2
∞ ∞
−∞ −∞
= − ∫ ∫
Using inversion formulae of two dimensional Canonical SS-Transform, g t x
( )
, is( )
{ }( )
2 2 2 2
2 2 2 2
,
2 2
sin sin 2 , ..(4.2)
i a i a i d i d
t x s w
b b b b
g t x
i i s w
e e t x e e DCSST s w dsdw
b b b b
π π − − ∞ ∞ − −
−∞ −∞
= − ∫ ∫
Taking complex conjugates we get
( )
{ }( )
2 2 2 2
2 2 2 2
,
2 2
sin sin 2 , .
i a i a i d i d
t x s w
b b b b
g t x
i i s w
e e t x e e DCSST s w dsdw
b b b b
π π ∞ ∞
−∞ −∞
− −
= − ∫ ∫
( ) ( )
, ,f t x g t x dxdt
∞ ∞
−∞ −∞
∫ ∫
( ), 2 i 2 i 2i abt2 2i abx2 sin s sin w 2i db s2 2i dbw2{2 }( ), .
f t x dxdt e e t x e e DCSST s w dsdw
b b b b
π π
∞ ∞ ∞ ∞
−∞ −∞ −∞ −∞
− −
= ∫ ∫ − ∫ ∫
By changing the order of the integration
( ) ( )
{ }( )
2 2 2 2 ( )
2 2 2 2
, ,
2 2
= 2 ,
1 1
2 2 sin sin ,
2 2
i d i d i a i a
s w t x
b b b b
f t x g t x dxdt
i i
DCSST s w dsdw
b b
s w
ib ib e e t x e e f t x dxdt
b b
ib ib
π π
π π
π π
∞ ∞
−∞ −∞
∞ ∞
−∞ −∞
∞ ∞
−∞ −∞
− −
−
∫ ∫
∫ ∫
∫ ∫
( ) ( ), , =4 2 {2 }( ){, 2 }( ), ...4.3
f t x g t x dxdt π DCSST s w DCSST s w dsdw
∞ ∞ ∞ ∞
−∞ −∞
∫ ∫
−∞ −∞∫ ∫
Putting f t x
( )
, = g t x( )
, in 4.3 we get( ), 2 =4 2 {2 }( ), 2
f t x dxdt π DCSST s w dsdw
∞ ∞ ∞ ∞
−∞ −∞
∫ ∫
−∞ −∞∫ ∫
______________________________________________________________________________
5 Properties of kernel
Kernels of 2D canonical SS- transform satisfied following property (1) ks1
( ) ( )
t s k, s2 x w, =ks1( ) ( )
s t k, s2 w x,if a = d (2) ks1
( ) ( )
t s k, s2 x w, ≠ks1( ) ( )
s t k, s2 w x, if a≠d(3) ks1
(
−t s k,) ( )
s2 x w, =ks1( ) (
t s k, s2 −x w,)
(4) ks1
(
− −t, s k) ( )
s2 x w, =ks1( ) (
t s k, s2 − −x, w)
Above properties of kernel are simple to prove.
CONCLUSION
In this paper 2D canonical SS- transform is generalized in the distributional sense. The analyticity and parsevals identity are proved, and also properties of kernel are mentioned.
REFERENCES
[1] Almeida Lufs B., “The fractional Fourier transform and time frequency,” representation IEEE trans. On Signal Processing, Vol. 42, No. 11, Nov. 1994.
[2] Bracewell R.N., “The Hartley Transform”, Oxford U.K. Oxford Univ. Press, 1986.
[3] Chavhan.S.B. and Borkar.V.C., Two dimensional generalized canonical cosine-cosine transform IJMES. Volume 1Issue 4 April (2012).
[4] D.Mendlovic and H.M.Ozactas, Fractional Fourier transform and their optical Implementation-I J.OPT.Soc.Am.A, 10, 1875-1881(1993).
[5] H.M.Ozactas and David Mendlovic, “Fourier transforms of Fractional order. and their optical interpitation,opt.commun.,101,163-169(1993).
[6] K.B.Walf, “Integral transforms in Science and engineering,” ch.9, New York plenum press, 1979.
[7] Zemanian A.H., “Distribution theory and transform analysis,” McGraw Hill, New York, 1965.
[8] Zemanian A.H., Generalized integral transform,” Inter Science Publisher’s New York, 1968.