Image transforms on a flourish¶

In [19]:

from numpy import *
from importlib_resources import read_binary
from io import BytesIO

In [20]:

import skimage.io, skimage.transform, skimage.exposure, skimage.morphology, skimage.filters, skimage.feature

In [21]:

flourish = skimage.io.imread(BytesIO(read_binary('deathbeds.data', 'IMG-0762.JPG')))
flat = skimage.color.rgb2gray(flourish)

The original drawing.¶

It was drawn on one of them big post its.

In [23]:

%matplotlib inline
skimage.io.imshow(flourish)

Out[23]:

<matplotlib.image.AxesImage at 0x1135e24a8>

In [24]:

%matplotlib inline
labelled =  skimage.measure.label(
        flat>.6)
skimage.io.imshow(labelled)

/Users/tonyfast/anaconda/envs/p6/lib/python3.6/site-packages/skimage/io/_plugins/matplotlib_plugin.py:74: UserWarning: Low image data range; displaying image with stretched contrast.
  warn("Low image data range; displaying image with "

Out[24]:

<matplotlib.image.AxesImage at 0x1138b5320>

In [26]:

%matplotlib inline

skimage.io.imshow(
    (1 - arange(labelled.shape[1])/labelled.shape[1]) * flat + 
    arange(labelled.shape[1])/labelled.shape[1] * labelled/labelled.max())

Out[26]:

<matplotlib.image.AxesImage at 0x1c275c55c0>

In [27]:

%matplotlib inline
skimage.io.imshow(
    skimage.measure.label(
        skimage.color.rgb2gray(flourish)>.6));

/Users/tonyfast/anaconda/envs/p6/lib/python3.6/site-packages/skimage/io/_plugins/matplotlib_plugin.py:74: UserWarning: Low image data range; displaying image with stretched contrast.
  warn("Low image data range; displaying image with "

Skeletonized flourish¶

In [28]:

%matplotlib inline
skimage.io.imshow(
    skimage.morphology.binary_dilation(
        skimage.morphology.skeletonize(skimage.color.rgb2gray(flourish)>.6), 
        skimage.morphology.disk(5)
    ));