Has anyone else noticed that the Google Cloud Vision OCR that processes the text in images operates starting top to bottom, then left to right for English? And that it didn’t used to?
The problem with this is generally:
We write in English like this.
So we want to read the lines from left to right, top to bottom.
We write
do english
not like
really this
Top to bottom, left to right. Which is how you’re reading it.
To report an issue, it is better to do it at Issue Tracker, rather than write a post at this forum.
You can submit your report here:
Issue reports> > Google reviews every new issue report submitted by users. Sometimes one of our staff will ask for clarification or followup. After we’re able to replicate the issue, we’ll tell you that it’s been forwarded to the appropriate team.> > Depending on the circumstances, we may be able to provide periodic updates while an issue is being looked at, but usually we cannot provide too many specifics about the exact cause of an issue, or when it will be fixed.> > When we’ve fixed an issue in production, we’ll indicate this and then we’ll close the issue.
See also: