persistence_transform.py

"""
========================================
Persistence Transform in Mapcubes
========================================

Persistence Transform is a simple image processing technique that is useful for the visualization
and depiction of gradually evolving structures.
This example illustrates persistence transform using the AIA_193_CUTOUT series.
"""

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.colors as colors
import matplotlib.cm as cm

import sunpy.map
import sunpy.physics.differential_rotation as diffrot
from sunpy.data.sample import AIA_193_CUTOUT01_IMAGE, AIA_193_CUTOUT02_IMAGE, AIA_193_CUTOUT03_IMAGE

###############################################################################
# We create the MapCube using the AIA_193_CUTOUT sample data.
# To create a MapCube, we can call Map directly but add in a keyword to output a MapCube instead.
aiamapcube = sunpy.map.Map(AIA_193_CUTOUT01_IMAGE, AIA_193_CUTOUT02_IMAGE,
                           AIA_193_CUTOUT03_IMAGE, cube=True)

###############################################################################
# For a data set consisting of N images with intensity values I(x,y,t), the Persistence Map Pn is a
# function of several arguments, namely intensity, location and time:
# Pn(x,y,tn)=Q(I(x,y,t≤tn))
persistence_maps = []
for i, map_i in enumerate(aiamapcube[1:]):
    per = np.array([aiamapcube[n].data for n in [i, i+1]]).max(axis=0)
    smap = sunpy.map.Map(per, map_i.meta)
    smap.plot_settings['cmap'] = cm.get_cmap('Greys_r')
    smap.plot_settings['norm'] = colors.LogNorm(100, smap.max())
    persistence_maps.append(smap)

###############################################################################
# This plots the original mapcube
aiamapcube.peek()

###############################################################################
# This plots the final mapcube after implementing the persistence transform
result_mapcube = sunpy.map.MapCube(persistence_maps)
result_mapcube.peek()
plt.show()