OCR accuracy when reading PDF



I am reading via OCR a value in a PDF file that is “L00” but OCR returns “LOO”. Any chance to correct this?



Can you check with number datatype instead of String?


Hi @sharad_kumar, OCR result can only be stored in a string variable.


what the result should look like?

L zero zero ?


Yes, L zero zero


ohh…i didnt work on it…Thanks for this information if i want to extract data from pdf which contains numbers what should we do…that time also we have to take String only??



in your case I can recommend you to create a “postprocessing” step - Replace Text action o -> 0


Yes, that’s what I did, but I thought there would be some other solution.


more intellectual OCR tuning and post-processing is available in the SPA product.


You extract as string and then convert that variable into number via clipboard operation for example. Make sure the extracted string variable does not contain other characters than numbers before conversion and trim whitespaces.


Thanks For this…