Note that the commands fft2, visualfft and conv require matrixes to have the names real, imag, spec and angle. And all matrixes have to be stored under a single data frame (which is an argument to the commands below).
| fft | Perform a one-dimensional Fourier transformation for a matrix |
| -d <data> | source data (contains "real" [and "imag"] matrix) |
| -dout <dataout> | target data |
| [-spec] | [-angle] | flags to add spectrum and angle matrixes to target |
| [-inv] | inverse transform |
This command performs a one-dimensional Fourier transformation.
| fft2 | Perform a two-dimensional Fourier transformation for a matrix |
| -in <matrix> | matrix to transform |
| -out <matrixout> | matrix to output |
| [-inv] | inverse transform |
| [-nulls <0 | 1 | 2>] | 0 = nulls for the left and top sides of the larger matrix |
| 1 = nulls for all the sides | |
| 2 = nulls for the right and bottom of the matrix (default) | |
| [-width <w>] | width for a larger matrix (rounded to the next integer power of 2) |
| [-height <h>] | height for a larger matrix (rounded to the next integer power of 2) |
| [-re, -im, -spec, -angle] | defines for the output matrixes (-re and -im are the default) |
| [-all] | create all output matrixes |
This command performs a two-dimensional Fourier transformation for matrixes.
| visualfft | Visualize a frame containing real and imag matrixes |
| -in <frame> | frame containing the real and imag matrixes |
| -out <image> | resulting image |
| -width <w> | resulting image width (must be <= frame matrixes) |
| -height <h> | resulting image height ( - " - ) |
| [-nulls <0 | 1 | 2>] | zero padding used in the fft-transform (see fft2) for cutting away blank areas (default 2) |
| [-shift] | use a shifting process for the frame |
| [-scale] | fit values to be shown in 256 gray levels |
This command visualizes a frame containing real and imag matrixes (usually after an fft-transform).
| conv | Perform a convolution between two frames |
| -in1 <frame1> | the first frame for convolution |
| -in2 <frame2> | the second frame for convolution |
| -out <oframe> | resulting frame |
This command makes a convolution between two frames (both including at least a real matrix).
Example (ex6.1):
... # 2-dimensional fft-transform NDA> fft2 -in mask -out maskfft -width 128 -height 128 NDA> fft2 -in matrix -out matrixfft # Convolution and visualization of the result NDA> conv -in1 matrixfft -in2 mask -out convmatrix NDA> visualfft -in convmatrix -out convimage -width 128 -height 128 -scale ...