In [1]:

    

%matplotlib inline


    

In [2]:

    

import numpy as np
import pylab
import mahotas as mh


    

In [3]:

    

img=mh.imread('Data/sample.tif')


    

In [4]:

    

pylab.imshow(img)


    

    
Out[4]:





<matplotlib.image.AxesImage at 0x7f6671233b90>






    

In [5]:

    

print img.shape


    

    




(233, 270, 4)

In [6]:

    

img=img[:,:,0]


    

In [20]:

    

print img.shape
print img.max()
print img.min()


    

    




(233, 270)
255
0

In [8]:

    

T = mh.thresholding.otsu(img)
pylab.imshow(img > T)


    

    
Out[8]:





<matplotlib.image.AxesImage at 0x7f66711368d0>






    

In [58]:

    

imgf = mh.gaussian_filter(img,4)
print imgf.shape
print imgf.max()
print imgf.min()


    

    




(233, 270)
192.60795559
75.8079078432

In [56]:

    

T= mh.thresholding.otsu(imgf.astype(np.uint8))
pylab.imshow(imgf > T)


    

    
Out[56]:





<matplotlib.image.AxesImage at 0x7f666ada3cd0>






    

In [59]:

    

rmax = mh.regmax(imgf)
se = np.ones((5,5))
rmax_big = mh.morph.dilate(rmax,se)
over = mh.overlay(imgf.astype(np.uint8),rmax)
over2= mh.overlay(imgf.astype(np.uint8),rmax_big)


    

In [60]:

    

pylab.imshow(over2)


    

    
Out[60]:





<matplotlib.image.AxesImage at 0x7f666b1fcc90>






    

In [29]:

    

seeds,nr_nuclei = mh.label(rmax)
print nr_nuclei


    

    




352

In [27]:

    

dist = mh.distance(imgf > T)
dist = dist.max() - dist
dist -= dist.min()
dist = dist/float(dist.ptp()) * 255
dist = dist.astype(np.uint8)
pylab.imshow(dist)


    

    
Out[27]:





<matplotlib.image.AxesImage at 0x7f666b51bfd0>






    

In [31]:

    

nuclei = mh.cwatershed(dist,seeds)
pylab.imshow(nuclei)


    

    
Out[31]:





<matplotlib.image.AxesImage at 0x7f666b4bc650>






    

In [33]:

    

whole = mh.segmentation.gvoronoi(nuclei)
pylab.imshow(whole)


    

    
Out[33]:





<matplotlib.image.AxesImage at 0x7f666b3f2550>






    

In [34]:

    

borders = np.zeros(nuclei.shape, np.bool)
borders[ 0,:] = 1
borders[-1,:] = 1
borders[:, 0] = 1
borders[:,-1] = 1
at_border = np.unique(nuclei[borders])
for obj in at_border:
    whole[whole == obj] = 0
pylab.imshow(whole)


    

    
Out[34]:





<matplotlib.image.AxesImage at 0x7f666b34c690>






    

In [38]:

    

print whole.shape
print whole.min()
print whole.max()


    

    




(233, 270)
0
339

In [42]:

    

se = np.ones((5,5))
rmax_big = mh.morph.dilate(rmax,se)


    

In [43]:

    

over = mh.overlay(imgf.astype(np.uint8),rmax_big)
pylab.imshow(over)


    

    
Out[43]:





<matplotlib.image.AxesImage at 0x7f666b0e60d0>






    

In [45]:

    

over2 = mh.overlay(img.astype(np.uint8),rmax_big)
mh.imsave('Data/copy.png',over2)


    

In [ ]: