In [1]:

    

from PIL import Image
from numpy import *
from pylab import *


    

In [2]:

    

import ncut
ncut = reload(ncut)
from scipy.misc import imresize


    

In [3]:

    

im = array(Image.open('C-uniform03.ppm'))
m, n = im.shape[:2]


    

In [4]:

    

wid = 50
rim = imresize(im, (wid, wid), interp='bilinear')
rim = array(rim, 'f')


    

In [5]:

    

A = ncut.ncut_graph_matrix(rim, sigma_d=1, sigma_g=1e-2)
code, V = ncut.cluster(A, k=3, ndim=3)
codeim = imresize(code.reshape(wid, wid), (m, n), interp='nearest')


    

In [6]:

    

figure()
imshow(codeim)
gray()
show()


    

In [7]:

    

code, V = ncut.cluster(A, k=4, ndim=3)
figure()
for i in range(4):
    subplot(1, 4, i+1)
    imshow(imresize(V[i].reshape(wid, wid), (m, n), interp='bilinear'))
    axis('off')
show()


    

In [8]:

    

def normal_cut(im, sd, sg):
    A = ncut.ncut_graph_matrix(rim, sigma_d=sd, sigma_g=sg)
    code, V = ncut.cluster(A, k=3, ndim=3)
    codeim = imresize(code.reshape(wid, wid), (m, n), interp='nearest')

    figure()
    gray()
    subplot(1, 2, 1)
    imshow(rim)
    subplot(1, 2, 2)
    imshow(codeim)
    show()

    code, V = ncut.cluster(A, k=4, ndim=3)
    figure()
    for i in range(4):
        subplot(1, 4, i+1)
        imshow(imresize(V[i].reshape(wid, wid), (m, n), interp='bilinear'))
        axis('off')
    show()


    

In [10]:

    

normal_cut(rim, 2, 1e-2)


    

In [11]:

    

normal_cut(rim, 0.5, 1e-2)


    

In [12]:

    

normal_cut(rim, 1, 2e-2)


    

In [13]:

    

normal_cut(rim, 1, 0.5e-2)


    

In [14]:

    

normal_cut(rim, 2, 2e-2)


    

In [15]:

    

normal_cut(rim, 0.5, 0.5e-2)


    

In [16]:

    

normal_cut(rim, 2, 0.5e-2)


    

In [17]:

    

# I like this one


    

In [ ]: