Configure OCR

In the following examples, OCR is configured to process scanned pages that contain only English or only Japanese text. Providing information about the input can result in a performance improvement, but OCR may fail to recognize text that does not match your configuration. For more information about optimizing OCR, see Optimize OCR Performance.

To configure OCR

  • To configure OCR through the Python API, call the method ocr on your session configuration. For example:

    Copy
    import keyview.filter as kv
    session = kv.FilterSession(bin_path, license)

    session.config.ocr(
        True,
        orientation=kv.OCROrientation.Upright,
        text_finding_mode=kv.OCRTextFindingMode.Document,
        detect_alphabet=kv.OCRDetectAlphabet.Listed,
        languages=["en", "ja"]
    )