Open
Description
The google OCR vision text API do not return the confidence level related to the recognized textblock.
This information would be useful to drop textblocks with low confidence level.
An issue I found is that characters like £ & ( ) and others are sometime wrongly recognized as letters or numbers...