Not 100% match in text search

Hi all you brilliant boys and girls out there!

We have a datable with company names. We also have an input of OCR text from a pdf (because it’s an image file). Problem with OCR is that the interpretation of characters are not always good. Depending on the image quality etc. Currently we are looping the datable and checking the if the OCR text contains the name of the datatable company. In most cases we will find a match. But when there are slight differences in the company name and the OCR we will not receive a match. The difference can be anywhere in the word. An example would be company name in table: “Craft Inc” But in the OCR text it appears as “Craff Inc”. Any good suggestions on how to solve this matching?

did we try with READ PDF with OCR activity and get the output with a variable of type string, and if used set the scale level to appropriate value so that we will be getting the right value
we can use CV activity if its a specific value from a PDF
we can use other OCR like abby flexicapture but we need license for that…

Cheers @RayJay


Actually your answer pointed us in the correct direction. We looked at the settings of the OCR again and found that the profile was set to “Legacy” instead of “Scan”. The results improved when we switched to “Scan” profile.
Thanks! :slight_smile:

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.