Regional recognition using zones, anchors, or the "Read field again" option normally performs OCR on each region individually.
Since Microsoft Vision charges per request, BIC always submits the full-page image to avoid multiple requests for documents having several zones on the same page.
BIC then saves the page results in the workdoc and reuses these results whenever possible.
- Due to non-zero rotation values, it is generally not possible to reuse page results saved during document recognition.
- Any differences in preprocessing settings for zones on the same page will most likely prevent the reuse of page results, and so requiring separate requests for each such zone
Further situations where CloudOCR results are not reused
- Testing zones in Designer
- Learning anchors in Designer
Note: Testing zones and learning anchors in Designer does not increment the CloudOCR
engine license counters.
For more information, see the Brainware Intelligent Capture Product Licensing Guide.