PyPI page
Home page
Author:
Igor Saevets
Summary:
Read a scanned page with several independent OCR engines and accept a word only when they agree. No language model in the transcription path.
Latest version:
0.2.2
Required dependencies:
numpy
|
onnxruntime
|
pillow
|
pymupdf
|
rapidocr
Optional dependencies:
numpy
|
onnxruntime
|
pillow
|
pymupdf
|
rapidocr
Downloads last day:
1
Downloads last week:
11
Downloads last month:
28
7-day smoothing