First implementation of the WavClumps algorithm, adapted to match the interface of acalib's algorithms.
It still relies on Matlab engine to perform the MRA with wavelets, so it's a requirement to have Matlab installed and configured to make it work.
In the near future, it would be desirable to remove this Matlab dependency by using PyWavelet python package https://pywavelets.readthedocs.io/en/latest/ . The adaptation is not straightforward because of the lack of wavedec3() and waverec3() functions in PyWavelets. The similar functions in this package are named wavedecn() and waverecn()http://pywavelets.readthedocs.io/en/latest/ref/nd-dwt-and-idwt.html, but they don't have the option to perform reconstruction from approximations or details (as Matlab does), but always perform perfect reconstruction (as I understand...)
First implementation of the WavClumps algorithm, adapted to match the interface of acalib's algorithms.
It still relies on Matlab engine to perform the MRA with wavelets, so it's a requirement to have Matlab installed and configured to make it work.
In the near future, it would be desirable to remove this Matlab dependency by using PyWavelet python package https://pywavelets.readthedocs.io/en/latest/ . The adaptation is not straightforward because of the lack of
wavedec3()
andwaverec3()
functions in PyWavelets. The similar functions in this package are namedwavedecn()
andwaverecn()
http://pywavelets.readthedocs.io/en/latest/ref/nd-dwt-and-idwt.html, but they don't have the option to perform reconstruction from approximations or details (as Matlab does), but always perform perfect reconstruction (as I understand...)