

In [26]:

    
from pylab import *
import ocrolib
from scipy.ndimage import filters,interpolation
from scipy import stats
from ocrolib import lineest
reload(lineest)









    Out[26]:





<module 'ocrolib.lineest' from 'ocrolib/lineest.pyc'>

Generate Distorted Image



In [27]:

    
figsize(8,8)
image = 1-ocrolib.read_image_gray("010036.bin.png")
image = interpolation.affine_transform(image,array([[1,0.03],[0,1]]),offset=(-30,0),output_shape=(100,700),order=0)
imshow(image,cmap=cm.gray)
h,w = image.shape

MvNormalizer



In [28]:

    
reload(lineest)
mv = lineest.MvNormalizer()
mv.measure(image)



In [29]:

    
imshow(mv.normalize(image,order=0),cmap=cm.gray)









    Out[29]:





<matplotlib.image.AxesImage at 0x607ae50>

Lineest Normalizer



In [30]:

    
reload(lineest)
le = lineest.load_normalizer("en-mixed.lineest")
le









    Out[30]:





<ocrolib.lineest.LineestNormalizer instance at 0x5ed8050>



In [31]:

    
le.measure(image)



In [32]:

    
imshow(le.normalize(image),cmap=cm.gray,interpolation='nearest')









    Out[32]:





<matplotlib.image.AxesImage at 0x66fbf50>



In [33]:

    
imshow(le.normalize(image,order=0)[:,:40],cmap=cm.gray,interpolation='nearest')









    Out[33]:





<matplotlib.image.AxesImage at 0x6bf08d0>

Saved MvNormalizer



In [34]:

    
reload(lineest)
le = lineest.load_normalizer("../models/mvnorm-24.lnorm")
le.measure(image)
imshow(le.normalize(image),cmap=cm.gray)
le.target_height









    Out[34]:





24

Saved Gradnorm (NB: no deskewing)



In [35]:

    
reload(lineest)
le = lineest.load_normalizer("../models/gradnorm-48.lnorm")
le.measure(image)
imshow(le.normalize(image),cmap=cm.gray)
le.target_height









    Out[35]:





48



In [35]: