text correction ?

[Wurschtbrot]

I made a OCR engine, but i have problems with 'I' and 'l' they have the same icon(pixels).

i have to compare a text-string with another string in a Dict and select best match is there a already a program ? is there a web based dict i can use ? maybe a free db ? or anything i can use ?

ATM i write all 'l' like a 'I' no casesesety atm only need the text overall... so L = I too

if this text is on screen

'eVery senTence Looks lIkE thIS'

becomes

'every sentence iooks iike this'

after OCR, i want this to be corrected to this

'every sentence looks like this'

that would be perfect

[rudi]

Hi.

[looking for a spell checker dict]

Spell Checking is a quite complex thing. You might want to have a look upon the poor spell checking for OpenOffice 2.x German Language: Try your pseudonym, Wurschtbrot, I doubt it will be corrected properly

As you know, especially in German there are tons of word combinations commonly used. Much, much more, than in most other languages. So either your dict will have to know how such combinations are allowed, or you will need to blow it to enormous size.

miXEd upper / lowercase you could catch using regex, searching for capital letters at "not-first-letter-positions".

Viel Glück, Rudi.