Skip to content

OCR-D workflow "Best results for selected pages" fails to detect text regions #451

@stweil

Description

@stweil

The recommended workflow does not recognize any text.
The dewarp step reports that there are no text regions for any page.
So it looks like the previous segmentation step is wrong:

[...]
19:08:24.661 INFO ocrd.task_sequence.run_tasks - Start processing task 'cis-ocropy-dewarp -I OCR-D-SEG -O OCR-D-SEG-LINE-RESEG-DEWARP -p '{"dpi": 0, "range": 4.0, "smoothness": 1.0, "max_neighbour": 0.05}''
Page "FILE_0001_OCR-D-SEG" contains no text regions
Page "FILE_0002_OCR-D-SEG" contains no text regions
Page "FILE_0003_OCR-D-SEG" contains no text regions
Page "FILE_0004_OCR-D-SEG" contains no text regions
Page "FILE_0005_OCR-D-SEG" contains no text regions
Page "FILE_0006_OCR-D-SEG" contains no text regions
[...]

Because of the missing text regions, the final text recognition terminates rather fast without any recognized text.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions