Ocr not able to parse numbers in this image


I am not able to get the yellow numbers in this image:

Anyone know why?


I confirmed that the OCR API has an issue with these numbers right now. The reason is that they are very “edgy” - they dont look like the normal numbers you find in a text. => Can you smoothen them somehow (remove edges)?

No, I am not able to smoothen them out.
Will you try to make the ocr api recognize numbers which are more “edgy”? :slight_smile:


This is on our todo list :slight_smile: