0

I swear I read somewhere that v1beta3 of the OCR Processor could do table extraction. I tested using the 2020 and 2022 version and have had no table data being returned. Did I just imagine what I read or am I doing something wrong?

Adam B
  • 95
  • 11

1 Answers1

1

For detecting and extracting data from tables, use the Form Parser processor

And it's recommended to use the v1 API for production applications rather than v1beta3 or v1beta2

Which Processor should I use?

Handle the processing response - Forms and Tables

General Processors Videos & Labs

Holt Skinner
  • 1,692
  • 1
  • 8
  • 21
  • I always wondered what service version was the latest - to me, v1beta3 could read as the beta 3 *before* v1, making the betas obsolete. Since there are still around, it's different. Is there public doc on what's different between the versions ? – yan-hic Nov 07 '22 at 20:40
  • 1
    You can check out the REST API references for the methods/resources for each version https://cloud.google.com/document-ai/docs/reference/rest `v1` is the Generally Available version, whereas `v1beta3` is a Preview version. Some newer features might be pushed to `v1beta3` before `v1`. However, I agree that the naming convention has been confusing because `v1beta3` did come out before `v1` initially. – Holt Skinner Nov 10 '22 at 16:44