MATLAB CVIP Toolbox Functions
Transform
The transform category contains the Fourier, cosine, Haar, Walsh-Hadamard and wavelet transforms as well as their inverse transform functions. For the Fourier transform, functions to extract either the magnitude or phase in 8-bit format are also included, so they can be displayed as images. When displaying transform spectra we typically use one of the mapping functions to remap the data, with the log remap being the most commonly used for spectral display.
The transforms can be performed on the entire image or on a block (subimage) by block basis by specification of the blocksize; or, in the case of the wavelet to specify the number of decomposition levels. As in CVIPtools, the Fourier spectrum has the origin shifted to the center, while the other transform functions have the origin in the upper left corner.
- dct_cvip - perform block-wise discrete cosine transform
- fft_cvip - performs fast Fourier transform
- fft_mag_cvip - extract magnitude of Fourier spectrum
- fft_phase_cvip - extract phase of Fourier spectrum
- haar_cvip - perform forward Haar transform
- idct_cvip - inverse discrete cosine transform
- ifft_cvip - performs inverse Fourier transform
- ihaar_cvip - perform inverse Haar transform
- iwalhad_cvip - perform inverse Walsh/Hadamard transform
- iwavdaub4_cvip - perform inverse wavelet transform based on Daubechies wavelet
- iwavhaar_cvip - perform inverse wavelet transform based on Haar wavelet
- wavdaub4_cvip - perform forward wavelet transform based on Daubechies wavelet
- wavhaar_cvip - perform forward wavelet transform based on Haar wavelet
- walhad_cvip - perform forward Walsh/Hadamard transform