[2.7.0] - Release Notes
In this version, we have improved the PDF processing in the OCR engine. Also, we have improved the licensing backend, with High Availability, and the facephi-recognition library.
Added
- PDF: Reduce default threshold to 0.5 for better detection of text.
- PDF: Add optional parameter
pageNums. It is a list of pages numbers to be processed. The counting starts at 1. If not present all pages will be processed. - PDF: Ignore errors "page not parsed" on 2nd and consecutive pages.
- PDF: Add output
warnings. This field contains information about the pages which have not been processed. In the future, it may be extended to contain other information. - Update text-interpreter:
- PDF: Include fields on the model although they are not detected or empty.
- Update document-reader-sdk-templates
- PDF: Models IVA.json and RENTA.json process correctly multi-word fields.
- Update licensing backend v2.0.0.
- Add high availability.
- Update facephi-recognition v3.21.0