[Image-SIG] Need better quality from crop

[Bryan Fodness]

I am trying to extract data from a large number of images.  I am
trying to use pytesser, but I am not getting the desired results.  I
tried to recognize the entire image and it was not consistent at all.
So, I cropped the picture with PIL in the sections that I am
interested in, and then tried to run.  The problem I am having is with
the date and time block as well as the numerical data.  I have tried
to use the ImageEnhance and ImageFilter but I have not been able to
increase the accuracy.  Does anyone have any experience making the
'text' in an image easily detectable for an ocr.

    import ImageEnhance, ImageFilter
    from pytesser import *

    im = Image.open('C:\\Users\\bryan\\Desktop\\10-28-08.bmp')

    im1 = im.crop([156,105,265,120])    # Date/Time
    im1 = im1.convert('RGB')
    im2 = im.crop([380,815,430,833])    # Couch Vrt
    im2 = im2.convert('RGB')

    text1, text2 = image_to_string(im1), image_to_string(im2)

>>> text1
'\n'
>>> text2
'HIS\n\n'
>>>

-- 
"The game of science can accurately be described as a never-ending
insult to human intelligence."	- João Magueijo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DateTime.bmp
Type: image/bmp
Size: 4974 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090219/2a9f51c1/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: CouchVrt.bmp
Type: image/bmp
Size: 2790 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090219/2a9f51c1/attachment-0001.bin>