Noodlesoft Forums

Following up from the other thread, here is another issue, that I am having

On some of receipts, that I am scanning there is a capital font, which is pretty widely spaced
And that leads to the OCR recognizing a blank space where there isn't one every once in a while

The actual word is PARKSCHEIN
Sometimes it might be recognized as PARKSC HEIN or PA RK SCHEIN or P A RKSCHEI N
You get the drift

Given that there are ten letters and thus nine space between letters, I am looking at 512 (if math hasn't let me down there) combinations of spaces/no spaces in that word. How to effectively deal with that?

If you can still isolate it into a custom attribute somehow (maybe surrounding text), you can use the replace text feature to remove the spaces. Otherwise, you'll need a script to handle this type of case.

Mr_Noodle wrote:If you can still isolate it into a custom attribute somehow (maybe surrounding text), you can use the replace text feature to remove the spaces. Otherwise, you'll need a script to handle this type of case.

Took me a minute to figure out what you meant... You mean the "replace text..." when applying the custom attribute in an action, right? That doesn't help, because I want to use it in a condition to check whether it falls into the categorie. Then I'll probably use a script action to check for the word where I can use regexs

Thx