LYZ-Li / eemd-matlab

EEMD family, Matlab version. PARFOR supported.

eemd-matlab

EEMD family, Matlab version. PARFOR supported.

run matlabpool open 8 to enable parallel computing.

core functions

emd.m

Empirical mode decomposition. No ensemble, no additional white noise.

c = emd(y, goal)

y: one-dimensional input data goal: target mode numbers to decompose

eemd.m

Ensemble empirical mode decomposition.

modes = eemd(y, goal, ens, nos)

Please note if you are running with a 64-bit Windows PC, Matlab will automatically run with eemd.mexw64 instead of this script.

eemd2.m

modes = eemd2(img, goal, ens, nos_wn)
[modes, G, D] = eemd2(img, goal, ens, nos_wn)

eemd3.m

modes = eemd3(img, goal, ens, nos_wn)
[modes, O, P, Q] = eemd3(img, goal, ens, nos_wn)

Demo files

eemd_2d_demo.m

Running this file will give you decomposed modes of a 64x64 Lena image.

eemd_3d_demo.m

Running this file will decompose a 54x54x54 volume of a Mimivirus. Please be carful as running this demo will cost up to 20 GB memory.

MEX compilaion

mex -IC:/path/to/boost_1_55_0 cpp/eemd.cpp

Replace C:/path/to/boost_1_55_0 with path to Boost on your computer.

Routine functions

zerox.m

find_extrema.m