volume_dwt

Perform forward or backward discrete wavelet transform on a MINC volume.

volume_dwt --forward <input> <output_LLL> ... <output_HHH>

volume_dwt --backward <input_LLL> ... <input_HHH> <output>

volume_dwt [--forward|--backward] --packed <input> <output>

DESCRIPTION

volume_dwt computes the discrete wavelet transform (DWT) on a 3D MINC volume. The forward transform decomposes the volume into eight subband components (LLL, LLH, LHL, LHH, HLL, HLH, HHL, HHH), where L denotes low-pass and H denotes high-pass filtering along each axis.

The inverse (backward) transform reconstructs the original volume from the eight subband components.

A packed mode is available where all subbands are stored in a single volume.

OPTIONS

--forward
Perform the forward discrete wavelet transform (decomposition).
--backward
Perform the backward (inverse) discrete wavelet transform (reconstruction).
--packed
Use packed mode where all subbands are stored in a single volume.

EXAMPLES

Forward DWT producing eight subband volumes:

volume_dwt --forward input.mnc LLL.mnc LLH.mnc LHL.mnc LHH.mnc HLL.mnc HLH.mnc HHL.mnc HHH.mnc

Inverse DWT reconstructing from subbands:

volume_dwt --backward LLL.mnc LLH.mnc LHL.mnc LHH.mnc HLL.mnc HLH.mnc HHL.mnc HHH.mnc output.mnc

Forward DWT in packed mode:

volume_dwt --forward --packed input.mnc packed_dwt.mnc

AUTHOR

Vladimir S. Fonov - McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University.

COPYRIGHTS

Copyright &copy; 2011 by Vladimir S. Fonov

SEE ALSO

fast_blur