Scribd
Upload
22 views4 pages

Jain SignalSystem

Uploaded by

mohammad zare
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
22 views4 pages

Jain SignalSystem

Uploaded by

mohammad zare
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 4
specify integer indices of arrays and vectors. The symbol roman j will represent ‘V=1. The complex conjugate of a complex variable such as z, will be denoted by 2*, Certain symbols will be redefined at appropriate places in the text to keep the notation cleat ‘Table 2.1 lists several well-known one-dimensional functions that will be often. ‘encountered. Their two-dimensional versions are functions of the separable form S59) =f) e) For example, the two-dimensional delta functions are defined as Dirac: 8s, ») = 8G)80) (22a) Kronecker: 8(m,1t)=8(n)8(1) (2.26) which satisfy the properties [free 00 =a'.y ye! dy = foe 9) tim f* ff 8G n9dedy = 1, 2.3) _Hemmd= SS at's a lon monn) an DE Som, m= 1 ‘The definitions and properties of the functions rect(sx, y), sine(x, y), and comb(x, y) can be defined in a similar manner. Sore Special Functions Function Definition Faetion Definition Recangle vents) {be EEY 1, x50 Signum sgn(a)=) 0, x =0 Sitting 1 r<0 property [7906 -2°) de . sinne saing Sine sine) = 8 property 6(ax)= 82 5 Keonect Comb comb(xy= & Ble -n) dela aay={ ™P0 1-bh, p= siting . Trangle ite) ~{h Ph EASY property DB f(mB(n~m)=F Chap.2 ‘Two-Dimensional Systems and Mathematical Preliminaries represent enoted by keep the IIbe often able form Qn (22a) (2.2) (2.4) combi, ») goss any vin Chap.2 2.3 LINEAR SYSTEMS AND SHIFT INVARIANCE A large number of imaging systems can be modeled as two-dimensional linear systems, Let x(m1t) and y(m, 1) represent the input and outpet sequences, respectively, of a two-dimensional system (Fig, 2.1), written as y(n) =# Em) e5) ‘This system is called linear if and only if any tinear combination of two inputs xm) and x, x) produces the same combination of their respective outputs y(n, a) and yalmn,n), ie., for arbitrary constants a, and ay a, H[x(m, a] + a: Lek, ny) ayu(onnt) + aya) (26) 2 [a,x(m, n) + aoxdon, n) This is called linear superpesition. When the input is the two-dimensional Kronecker delta function at location (va, 2"), the output at location (m,n) is defined as hem, nym, 2 VA H[B(m — mn nD 27) and is called the impuise response of the system. For an imaging system, itis the image in the output plane due to an ideal point source at location (ann) in the input plane, In our notation, the semicolon (;) is employed to distinguish the input and output pairs of coordinates. ‘The impulse response is called the pains spread function (PSF) when the inputs and outputs represent a positive quantity such as the intensity of light in imaging systems. The term impulse response is more general and is allowed to lake negative as well as complex values, The region of support of an impulse response is the smallest closed region in the m, » plane outside which the impulse response is zero. ‘A systems is said to be a finite impuise response (FIR) or an infinite impulse response (IR) system if its impulse response has finite or infinite regions of support, re- spectively. “The output of any linear system can he obtained from its impulse response and the input by applying the superposition rule of (2.6) to the representation of (2.4) as follows: ym, n) = H[x(m, =H|DLaxm', non — mn —n’) = LD ena x(a mn =n) D> yomn)= ES Dx6n' 0 adn, ayn’) (2.8) 2m 0) —+ [2] — vim) Figure 24. A system See.2.3 Linear Systems and Shift Invariance a A system is called spatially invariant ot shife invariant if a translation of the input causes « translation of the output, Following the definition at (2.7), if the impulse ‘cours atthe origin we will have S{o(0m,n)] = (om, 50,0) Hence, it must be true for shift invariant systems that ilo, ropa’, n') 838 (n = mt’, = = h(n ~ m,n = 1":0,0) Sl mia'sa) = Alen msn =n) eo i.e., the impulse response is a function of the two displacement variables only. This ‘means the shape of the impulse response does not change as the impulse moves about the m,n plane. A system is called spatially varying when (2.9) does not hold. Figure 2.2 shows examples of PSFs of imaging systems with separable or circularly syinmetric impulse responses, For shift invariant aystems, the output becomes ylon.n)= SS naam’. —w)xton'sn") 2.19 Which is called the convolution of the input with the impulse response. Figure 2.3 shows a graphical intexpretation of this operation. The impulse response array is rotated about the origin by 180° and then shifted by (m,n) and overlayed on the atray x(m‘,n"). The sum of the product of the arrays {x(-.)} and {(-.)} in the ‘overlapping regions gives the result at (m,n). We will use the symbol @ to denote the convolution operation in both discrete and continuous eases Genoff nee neiy rye y de ay! him, n)Ox(m,n) 4 TY Adm mijn =n \x(m'n') (LAY ay) Bo igure 2.2. Examples of PSFe[ © 14 (2) Ciculety spmmetic PSF of average atmosphere turbulence e2using smal blur; (b) atmospheric tusbulease PSF ‘causing lage blur; () separable PSF of difracton ited system with square aerate; (6) same a (e) But with smaller aperture ‘Two-Dimensional Systems and Mathematical Preliminaries Chap. 2 of the input the impulse 29) es only. This pulse moves yes not hold. or cireulacly (2.20) - Figure 2.3 nse array is layed on the Go) im the @ to denote “ay! nw’) 2.11) ale psrsteld PSF of average ausing small, balence PSF nrable PSF of « ith square Phat with Chap. 2 Roteey 180° aa st by, sion’) (alimpatee response {ouput oestion tm} ethe sum of rode, el quantities ne area fovea, Figure 2.3. Disercte convolution in rwo dimensions ‘The convolution operation has several interesting properties, which are explored in Problems 2.2 and 2.3, Example 2.1 (Discrete convolution) Consider she 2 2 and 3» 2 arays hom} and x(n) shown ext, where the boxed lemeat is tthe origi, Alo shown ate the vriows steps for obtaining the convolution ofthese two arrays, The result (mn) iva4%3 array. In goneral. the comolution of two arrays of sizes (Mi Ni) and (Ms Na) yes am array of size [0M Ma 1) (N+ N= 1] Problem 2.5). tabs, Joep to a (x(a) (b) Adm, 0) ()AGom,

You might also like