1

I have never used python before, and I am not sure where to start. My goal is to take image data, of numbers and multicolored background, and reliably get the correct characters identified. I looked into the tools necessary for this and I found the Anaconda python distribution which included all the possible packages I might need for this, as well as tesseract-ocr and pytesser.

Unfortunately, I'm lost in how to begin. I"m using the PyCharm Community IDE and simply trying to follow this guide: http://www.manejandodatos.es/2014/11/ocr-python-easy/ to get a grasp on OCR.

This is the code I'm using:

from PIL import Image
from pytesser import *

image_file = 'menu.jpg'
im = Image.open(image_file)
text = image_to_string(im)
text = image_file_to_string(image_file)
text = image_file_to_string(image_file, graceful_errors=True)
print "=====output=======\n"
print text

and I believe the Anaconda distribution that I'm using has PIL, but I'm getting this error:

C:\Users\diego_000\Anaconda\python.exe C:/Users/diego_000/PycharmProjects/untitled/test.py
Traceback (most recent call last):
  File "C:/Users/diego_000/PycharmProjects/untitled/test.py", line 2, in <module>
    from pytesser import *
  File "C:\Users\diego_000\PycharmProjects\untitled\pytesser.py", line 6, in <module>
    import Image
ImportError: No module named Image

Process finished with exit code 1

Can anyone point me in the right direction?

Anthon
  • 69,918
  • 32
  • 186
  • 246
Ryan
  • 47
  • 1
  • 1
  • 6
  • Install Pillow. Use `from PIL import Image` – Malik Brahimi Jun 28 '15 at 19:52
  • can you explain this a bit more? I already have that line written, so would I just install Pillow? Or would I delete the following line `from pytesser import *`? – Ryan Jun 28 '15 at 20:27

1 Answers1

2

The document you point to says to use

from PIL import Image

except you use 

import Image

and so the interpreter properly says:

ImportError: No module named Image

It looks as if you reordered the lines

from PIL import Image
from pytesser import *

and that pytesser has a improperly coded dependency on PIL. but I can't be certain with the code you provided.

msw
  • 42,753
  • 9
  • 87
  • 112
  • I replaced the line `import image` in pytesser.py with `from PIL import Image` and my code compiles, but I do not get any output, this must mean that the code is flawed in some way? I did replace the original .tiff image with a jpg form of it, would you happen to know if that could affect it at all? – Ryan Jun 28 '15 at 20:30
  • It could very well effect it. TIFF is an uncompressed file and will have cleaner letter edges than a compress JPG. – msw Jun 28 '15 at 20:38
  • Everything works fine when using pytesser's included fonts_test.png file, but there is a problem whenever I use any other image. I imagine it's just a limitation of pytesser then? – Ryan Jun 28 '15 at 20:50