: Is it possible to configure OCR software so that it knows characters will only be numbers i.e. 0 to 9 ?
Answer : No, it is not possible to have OCR software set to only read information as numbers or only as letters. There are two methods that OCR software does have to accomplish this goal though.
The first is OCR training, which consists of going through a process of telling the software that characters that are being read incorrectly are another character than the one that is reading. The more training you do the more likely it is to start reading a character properly that is has been reading incorrectly consistently. The software that includes this training capability is the standard OCR software. We recommend FineReader, Readiris or OmniPage.
The training process can be tedious and will take a while, but once completed then you will get much better results on images that were consistently reading the incorrect characters. If you don't want to have to go through the training process or you also need some additional indexing capabilities you can use the second method of character correction with document scanning and indexing software.
With document scanning software you can set the OCR to always replace certain characters with other characters. This means that you can pre-set certain characters that always read incorrectly to other characters without going through a software training process. For example you can have all O's replaced with 0's or all I's replaced with 1's. This software also allows you to have index fields that will automatically store the images in certain folders and filenames as well as having a searching capability built in. The document scanning and indexing software that we would recommend is SimpleIndex.