Multidimensional transform

In mathematical analysis and applications, multidimensional transforms are used to analyze the frequency content of signals in a domain of two or more dimensions.

Uses your system voice; voice and rate in settings.

Snapshot

≈ 9 min read

Names in other languages

No names in other languages are held for this entry yet.

Article bodies are English and are not machine-translated. A name in another language is not a translation of the article.

In mathematical analysis and applications, multidimensional transforms are used to analyze the frequency content of signals in a domain of two or more dimensions. F(w_1,w_2,\dots,w_m) = \sum_{n_1=-\infty}^\infty \sum_{n_2=-\infty}^\infty \cdots \sum_{n_m=-\infty}^\infty f(n_1,n_2,\dots,n_m) e^{-i w_1 n_1 -i w_2 n_2 \cdots -i w_m n_m} where F stands for the multidimensional Fourier transform, m stands for multidimensional dimension.

Define f as a multidimensional discrete-domain signal. The inverse multidimensional Fourier transform is given by f(n_1,n_2,\dots,n_m) = \left(\frac{1}{2 \pi}\right)^m \int_{- \pi}^{\pi} \cdots \int_{-\pi}^{\pi} F(w_1,w_2,\ldots,w_m) e^{i w_1 n_1 +i w_2 n_2 + \cdots+i w_m n_m} \, dw_1 \cdots \,dw_m The multidimensional Fourier transform for continuous-domain signals is defined as follows: F(\Omega_1,\Omega_2,\ldots,\Omega_m) = \int_{-\infty}^{\infty} \cdots \int_{-\infty}^{\infty} f(t_1,t_2,\ldots,t_m) e^{-i \Omega_1 t_1-i \Omega_2 t_2 \cdots -i \Omega_m t_m} \, dt_1 \cdots \,dt_m

Linearity

if x_1(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X_1(\omega_1,\ldots,\omega_M) , and x_2(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X_2(\omega_1,\ldots,\omega_M) then, a x_1(n_1,\ldots,n_M) + b x_2(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} a X_1(\omega_1,\ldots,\omega_M) + b X_2 (\omega_1, \ldots, \omega_M)

Shift

if x(n_1,...,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X(\omega_1,...,\omega_M) , then x(n_1 - a_1,...,n_M - a_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} e^{-i(\omega_1 a_1 +,...,+ \omega_M a_M)} X(\omega_1,...,\omega_M)

Modulation

if x(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X(\omega_1,\ldots,\omega_M) , then e^{i(\theta_1 n_1 +\cdots+ \theta_M n_M)} x(n_1 - a_1,\ldots,n_M - a_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X(\omega_1 - \theta_1,\ldots,\omega_M - \theta_M)

Multiplication

if x_1(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X_1(\omega_1,\ldots,\omega_M) , and x_2(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X_2 (\omega_1,\ldots,\omega_M) then, {{NumBlk|::| x_1(n_1,\ldots,n_M) x_2(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} \frac{1}{(2\pi)^M} \int \limits_{-\pi}^\pi \cdots \int\limits_{-\pi}^\pi X_1(\omega_1 - \theta_1,\ldots,\omega_M - \theta_M) X_2 (\theta_1, \ldots, \theta_M) \, d\theta_1 \cdots d\theta_M | }} or, {{NumBlk|::| x_1(n_1,\ldots,n_M) x_2(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} \frac{1}{(2\pi)^M} \int\limits_{-\pi}^\pi \cdots \int\limits_{-\pi}^\pi X_1(\theta_1,\ldots,\theta_M) X_2(\omega_1 - \theta_1,\ldots,\omega_M - \theta_M) \, d\theta_1\cdots d\theta_M | }}

Differentiation

If x(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X(\omega_1,\ldots,\omega_M) , then -in_1x(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} \frac{\partial}{(\partial\omega_1)} X(\omega_1,\ldots,\omega_M), -in_2x(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} \frac{\partial}{(\partial\omega_2)} X(\omega_1,\ldots,\omega_M), (-i)^M(n_1n_2\cdots n_M)x(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} \frac{(\partial)^M}{(\partial\omega_1\partial\omega_2\cdots\partial\omega_M)} X(\omega_1,\ldots,\omega_M),

Transposition

If x(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X(\omega_1,\ldots,\omega_M) , then x(n_M,\ldots,n_1) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X (\omega_M,\ldots,\omega_1)

Reflection

If x(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X (\omega_1,\ldots,\omega_M) , then x(\pm n_1,\ldots,\pm n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X(\pm \omega_1,\ldots,\pm \omega_M)

Complex conjugation

If x(n_1,\ldots,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X(\omega_1,\ldots,\omega_M) , then x^{*}(\pm n_1,\ldots,\pm n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X^{*}(\mp \omega_1,\ldots,\mp \omega_M)

Parseval's theorem (MD)

if x_1(n_1,...,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X_1(\omega_1,...,\omega_M) , and x_2(n_1,...,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X_2(\omega_1,...,\omega_M) then, \sum_{n_1=-\infty}^\infty ... \sum_{n_M =-\infty}^\infty x_1 (n_1,...,n_M) x_2^{*}(n_1,...,n_M) {=} \frac{1}{(2\pi)^M} \int\limits_{-\pi}^{\pi} ... \int\limits_{-\pi}^{\pi}X_1(\omega_1,...,\omega_M) X_2^{*}(\omega_1,...,\omega_M)d\omega_1...d\omega_M if x_1(n_1,...,n_M) {=} x_2(n_1,...,n_M) , then \sum_{n_1=-\infty}^\infty ... \sum_{n_M =-\infty}^\infty |x_1 (n_1,...,n_M)|^2 {=} \frac{1}{(2\pi)^M} \int\limits_{-\pi}^{\pi} ... \int\limits_{-\pi}^{\pi}|X_1(\omega_1,...,\omega_M)|^2 d\omega_1...d\omega_M A special case of the Parseval's theorem is when the two multi-dimensional signals are the same. In this case, the theorem portrays the energy conservation of the signal and the term in the summation or integral is the energy-density of the signal.

Separability

A signal or system is said to be separable if it can be expressed as a product of 1-D functions with different independent variables. This phenomenon allows computing the FT transform as a product of 1-D FTs instead of multi-dimensional FT. if x(n_1,...,n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} X(\omega_1,...,\omega_M) , a(n_1) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} A(\omega_1) , b(n_2) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} B(\omega_2) ... y(n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} Y(\omega_M) , and if x(n_1,...,n_M) {=} a(n_1)b(n_2)...y(n_M) , then X(\omega_1,...,\omega_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} x(n_1,...,n_M) {=} a(n_1)b(n_2)...y(n_M) \overset{\underset{\mathrm{FT}}{}}{\longleftrightarrow} A(\omega_1) B(\omega_2)...Y(\omega_M) , so X(\omega_1,...,\omega_M) {=} A(\omega_1) B(\omega_2)...Y(\omega_M) Fx(K_1,K_2,\ldots,K_m)= \sum_{n_1=0}^{N_1-1} \cdots \sum_{n_m=0}^{N_m-1} fx(n_1,n_2,\ldots,n_m) e^{-i \frac{2 \pi}{N_1} n_1 K_1 -i \frac{2 \pi}{N_2} n_2 K_2 \cdots -i \frac{2 \pi}{N_m} n_m K_m} for , .

The inverse multidimensional DFT equation is fx(n_1,n_2,\ldots,n_m)= \frac{1}{N_1 \cdots N_m} \sum_{K_1=0}^{N_1-1} \cdots \sum_{K_m=0}^{N_m-1} Fx(K_1,K_2, \ldots ,K_m) e^{i \frac{2 \pi}{N_1} n_1 K_1 +i \frac{2 \pi}{N_2} n_2 K_2\cdots+i \frac{2 \pi}{N_m} n_m K_m} for . Fx(K_1,K_2,\ldots,K_r ) = \sum_{n_1=0}^{N_1-1} \sum_{n_2=0}^{N_2-1} \cdots \sum_{n_r=0}^{N_r-1} fx(n_1,n_2,\ldots,n_r) \cos { \frac{ \pi (2n_1+1) K_1}{2N_1}} \cdots \cos { \frac{ \pi (2n_r+1) K_r}{2N_r}} for , i = 1, 2, ..., r. F(s_1,s_2,\ldots,s_n) = \int_{0}^{\infty} \cdots \int_{0}^{\infty} f(t_1,t_2,\ldots,t_n) e^{-s_nt_n -s_{n-1}t_{n-1} \cdots \cdots s_1t_1} \, dt_1 \cdots \,dt_n where F stands for the s-domain representation of the signal f(t). A special case (along 2 dimensions) of the multi-dimensional Laplace transform of function f(x,y) is defined as F(s_1,s_2)= \int\limits_{0}^{\infty}\int\limits_{0}^{\infty}\ f(x,y) e^{-s_1x-s_2y}\, dxdy F(s_1,s_2) is called the image of f(x,y) and f(x,y) is known as the original of F(s_1,s_2) .

This special case can be used to solve the Telegrapher's equations. F(z_1,z_2,\ldots,z_m)= \sum_{n_1=-\infty}^{\infty} \cdots \sum_{n_m=-\infty}^{\infty} f(n_1,n_2,\ldots,n_m) z_1^{-n_1} z_2^{-n_2} \ldots z_m^{-n_m} where F stands for the z-domain representation of the signal f(n). A special case of the multidimensional Z transform is the 2D Z transform which is given as F(z_1,z_2)= \sum_{n_1=-\infty}^{\infty} \sum_{n_2=-\infty}^{\infty} f(n_1,n_2) z_1^{-n_1} z_2^{-n_2} The Fourier transform is a special case of the Z transform evaluated along the unit circle (in 1D) and unit bi-circle (in 2D). i.e. at z=e^{iw} where z and w are vectors.

Points (z 1 ,z 2 ) for which F(z_1,z_2)=\sum_{n_1=-\infty}^\infty \sum_{n_2=-\infty}^\infty |f(n_1,n_2)| |z_1|^{-n_1} |z_2|^{-n_2} are located in the ROC. An example: If a sequence has a support as shown in Figure 1.1a, then its ROC is shown in Figure 1.1b. This follows that |F(z 1 ,z 2 )| (z_{01},z_{02}) lies in the ROC, then all points (z_1,z_2) that satisfy |z1|≥|z01| and |z2|≥|z02| lie in the ROC.

Therefore, for figure 1.1a and 1.1b, the ROC would be \ln|z_1| \ge \ln|z_{01}| \text{ and } \ln|z_2| \ge L \ln|z_1| + \{ \ln|z_{02}| - L\ln|z_{01}| \} where L is the slope. The 2D Z-transform, similar to the Z-transform, is used in multidimensional signal processing to relate a two-dimensional discrete-time signal to the complex frequency domain in which the 2D surface in 4D space that the Fourier transform lies on is known as the unit surface or unit bicircle.

Partial differential equations

Discrete Fourier transforms are often used to solve partial differential equations, where again the DFT is used as an approximation for the Fourier series (which is recovered in the limit of infinite N). The advantage of this approach is that it expands the signal in complex exponentials e inx , which are eigenfunctions of differentiation: d/dx e inx = in e inx . Thus, in the Fourier representation, differentiation is simple—we just multiply by i n. (Note, however, that the choice of n is not unique due to aliasing; for the method to be convergent, a choice similar to that in the trigonometric interpolation section above should be used.)

A linear differential equation with constant coefficients is transformed into an easily solvable algebraic equation. One then uses the inverse DFT to transform the result back into the ordinary spatial representation. Such an approach is called a spectral method.

DCTs are also widely employed in solving partial differential equations by spectral methods, where the different variants of the DCT correspond to slightly different even/odd boundary conditions at the two ends of the array. Laplace transforms are used to solve partial differential equations.

The general theory for obtaining solutions in this technique is developed by theorems on Laplace transform in n dimensions. The multidimensional Z transform can also be used to solve partial differential equations. w(x,y)=\frac{1}{4} \left(1 + \cos {\frac {x \pi} N }\right)\left(1 + \cos {\frac{y \pi} N }\right) where N is the image dimension and x, y are the coordinates from the center of image spans from 0 to N/2.

The author wanted to compute an equal value for spatial frequency such as: \begin{align} A_m(f)^2= \left[ \sum_{i=-f}^f \right. & \operatorname{FFT}(-f,i)^2+ \sum_{i=-f}^f \operatorname{FFT}(f,i)^2 \\[5pt] & \left. {} + \sum_{i=-f+1}^{f-1} \operatorname{FFT}(i,-f)^2+ \sum_{i=-f+1}^{f-1} \operatorname{FFT}(i,f)^2 \right] \end{align} where "FFT" denotes the fast Fourier transform, and f is the spatial frequency spans from 0 to . The proposed FFT-based imaging approach is diagnostic technology to ensure a long life and stable to culture arts.

This is a simple, cheap which can be used in museums without affecting their daily use. But this method doesn’t allow a quantitative measure of the corrosion rate.

Text derived from https://en.wikipedia.org/w/index.php?oldid=1377588893 under CC BY-SA 4.0; modifications have been made.

Sources cited in this revision

3 named source(s) extracted from the cited revision. A source with no author is shown by its publication; nothing is invented.