from PIL import  Image
 
img = Image.open('input.gif')
img = img.convert("RGBA")
 
pixdata = img.load()
 
# Clean the background noise, if color != black, then set to white.
for y in xrange(img.size[1]):
    for x in xrange(img.size[0]):
        if pixdata[x, y] != (0, 0, 0, 255):
            pixdata[x, y] = (255, 255, 255, 255)
 
img.save("input-black.gif", "GIF")
 
#   Make the image bigger (needed for OCR)
im_orig = Image.open('input-black.gif')
big = im_orig.resize((116, 56), Image.NEAREST)
 
ext = ".tif"
big.save("input-NEAREST" + ext)
 
#   Perform OCR using pytesser library
from pytesser import *
image = Image.open('input-NEAREST.tif')
print image_to_string(image)