In [1]:

    

import cv2


    

In [2]:

    

import urllib.request


    

In [3]:

    

import numpy as np


    

In [4]:

    

import glymur


    

In [5]:

    

#import matplotlib.pyplot as plt
from matplotlib import pyplot as plt


    

In [6]:

    

%matplotlib inline


    

In [7]:

    

HTTP_response = urllib.request.urlopen('http://www.hpca.ual.es/~vruiz/images/lena.png') # get the image


    

In [8]:

    

arr = np.asarray(bytearray(HTTP_response.read()), dtype=np.uint8) # convert the image to a numpy.ndarray


    

In [9]:

    

img = cv2.imdecode(arr,-1) # Decompress the image


    

In [10]:

    

img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) # Convert to BGR -> RGB


    

In [11]:

    

plt.imshow(img)


    

    
Out[11]:





<matplotlib.image.AxesImage at 0x7fc714294780>






    

In [12]:

    

buf = img.copy() # Create a copy of arr


    

In [13]:

    

gaussian = [img] # The base of the gaussian pyramid is buf


    

In [14]:

    

# Create the rest of the gaussian pyramid
for i in range(5):
    print(i, len(gaussian), len(gaussian[i]))
    buf = cv2.pyrDown(buf)
    gaussian.append(buf)


    

    




0 1 512
1 2 256
2 3 128
3 4 64
4 5 32

In [15]:

    

plt.imshow(gaussian[0]);


    

In [16]:

    

plt.imshow(gaussian[1])


    

    
Out[16]:





<matplotlib.image.AxesImage at 0x7fc7141e8c18>






    

In [17]:

    

plt.imshow(gaussian[2])


    

    
Out[17]:





<matplotlib.image.AxesImage at 0x7fc6f439f940>






    

In [18]:

    

plt.imshow(gaussian[3])


    

    
Out[18]:





<matplotlib.image.AxesImage at 0x7fc6f4387550>






    

In [19]:

    

plt.imshow(gaussian[4])


    

    
Out[19]:





<matplotlib.image.AxesImage at 0x7fc6f42ee1d0>






    

In [20]:

    

plt.imshow(gaussian[5])


    

    
Out[20]:





<matplotlib.image.AxesImage at 0x7fc6f42d2b70>






    

In [21]:

    

laplacian = [gaussian[5]] # The upper level of the pyramid


    

In [22]:

    

# Create the rest of the laplacian pyramid
for i in range(5):
    interpolation = cv2.pyrUp(gaussian[5-i])
    #difference = cv2.subtract(gaussian[5-i-1],interpolation)
    difference = gaussian[5-i-1] - interpolation + 128
    laplacian.append(difference)


    

In [45]:

    

print (type(gaussian[5]))


    

    




<class 'numpy.ndarray'>

In [24]:

    

plt.imshow(laplacian[5])


    

    
Out[24]:





<matplotlib.image.AxesImage at 0x7fc6f4242c50>






    

In [25]:

    

color = ('b','g','r')
for i,col in enumerate(color):
    histr = cv2.calcHist([laplacian[5]],[i],None,[256],[0,256])
    plt.plot(histr,color = col)
    plt.xlim([0,256])
plt.show()


    

In [26]:

    

plt.imshow(laplacian[4])


    

    
Out[26]:





<matplotlib.image.AxesImage at 0x7fc6f4091dd8>






    

In [27]:

    

plt.imshow(laplacian[3])


    

    
Out[27]:





<matplotlib.image.AxesImage at 0x7fc6eb7d77b8>






    

In [28]:

    

plt.imshow(laplacian[2])


    

    
Out[28]:





<matplotlib.image.AxesImage at 0x7fc6f40b8b38>






    

In [29]:

    

plt.imshow(laplacian[1])


    

    
Out[29]:





<matplotlib.image.AxesImage at 0x7fc6f40f9dd8>






    

In [30]:

    

plt.imshow(laplacian[0])


    

    
Out[30]:





<matplotlib.image.AxesImage at 0x7fc6eb7bda90>






    

In [31]:

    

# Reconstruction of the image using the laplacian pyramid
reconstruction = laplacian[0]
for i in range(5):
    reconstruction = cv2.pyrUp(reconstruction)
    reconstruction = reconstruction + laplacian[i+1] - 128
    #reconstruction = cv2.add(reconstruction, np.uint8(laplacian[i+1]/0.5))


    

In [32]:

    

plt.imshow(reconstruction)


    

    
Out[32]:





<matplotlib.image.AxesImage at 0x7fc6f448bb70>






    

In [33]:

    

color = ('b','g','r')
for i,col in enumerate(color):
    histr = cv2.calcHist([reconstruction],[i],None,[256],[0,256])
    plt.plot(histr,color = col)
    plt.xlim([0,256])
plt.show()


    

In [34]:

    

color = ('b','g','r')
for i,col in enumerate(color):
    histr = cv2.calcHist([img],[i],None,[256],[0,256])
    plt.plot(histr,color = col)
    plt.xlim([0,256])
plt.show()


    

In [35]:

    

jp2 = glymur.Jp2k('lena.jp2', data=img)


    

In [36]:

    

laplacian[0].shape[1]


    

    
Out[36]:





16

In [44]:

    

for i in range(5):
    #jp2 = glymur.jp2box.ImageHeaderBox(laplacian[i+1].shape[0], laplacian[i+1].shape[1], num_components=3, signed=True, bits_per_component=8, compression=7, colorspace_unknown=False, ip_provided=False, length=0, offset=-1)
    jp2 = glymur.Jp2k('laplacian_{}.jp2'.format(i+1), data=laplacian[i+1], numres=1)
    #jp2 = glymur.Jp2k('laplacian_{}.jp2'.format(i+1), mode='wb')
    #jp2.write(laplacian[i+1])
    #print (jp2)


    

In [ ]:

    




    

In [ ]:

    




    

In [ ]: