How can I extract a single component from a Fourier decomposition of a 2D matrix and plot it as a 2D map?

Question

Carlo 2015-3-24

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/184950-how-can-i-extract-a-single-component-from-a-fourier-decomposition-of-a-2d-matrix-and-plot-it-as-a-2d

评论： Rick Rosson 2015-3-26

在 MATLAB Online 中打开

I have a scalar field (potential vorticity PV) on a 2-D square grid (599x599 cells).

I need to do a Fourier decomposition of the field and extract the contribution of the most important components.

Then, I have to plot the map of each component alone.

I would like to know how to extract the single components of the field and plot them as 2D (599x599) maps.

So far:

    % Spatial sampling intervals 
    % (in km per pixel):
    dx = 15;
    dy = 15;  
    F1 = fft2(PV);  % 2D Fourier Transform of the matrix PV
    F2 = fftshift(F1);  % zero-frequency component in the middle of the spectrum
    imagesc(abs(F2)),colorbar

Not sure what to do after this.. any help would be appreciated!

2 个评论
显示无隐藏无

Rick Rosson 2015-3-24

编辑：Rick Rosson 2015-3-24

在 MATLAB Online 中打开

It would be helpful if you could provide the spatial sampling intervals in the x- and y directions (in units of distance per pixel, e.g. meters, cm, feet, or inches per pixel). For example:

   % Spatial sampling intervals 
   % (in meters per pixel):
   dx = 0.02;
   dy = 0.03;

These parameters would allow you to calibrate the spatial frequency in the Fourier domain, which you will need to do to determine which spectral components are most prominent.

Carlo 2015-3-24

Thanks Rick, I have added the spatial sampling intervals in the x- and y- directions

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Rick Rosson 2015-3-24

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/184950-how-can-i-extract-a-single-component-from-a-fourier-decomposition-of-a-2d-matrix-and-plot-it-as-a-2d#answer_172593

编辑：Rick Rosson 2015-3-24

在 MATLAB Online 中打开

Here's a start:

[ M, N ] = size(PV);
Fs_x = 1/dx;
Fs_y = 1/dy;
dF_x = Fs_x/M;
dF_y = Fs_y/N;
Fx = -Fs_x/2:dF_x:Fs_x/2-dF_x + mod(M,2)*dF_x/2;
Fy = -Fs_y/2:dF_y:Fs_y/2-dF_y + mod(N,2)*dF_y/2;
F = meshgrid(Fx,Fy);

5 个评论
显示 3更早的评论隐藏 3更早的评论

Carlo 2015-3-26

在 MATLAB Online 中打开

Do I need these wavenumber matrices? To display the spectral image in the wavenumber domain I am using Kx and Ky to represent the X- and Y- axes, respectively:

Kx = 2*pi*Fx;
Ky = 2*pi*Fy;
figure;pcolor(Kx,Ky,abs(F2)),shading flat,colorbar

I am still not sure about how to have a 599x599 map in the x,y domain with only a specific wavenumber component of the original map. Should I do some filtering on the 2D FDT map?

Rick Rosson 2015-3-26