1
0
Fork 0
mirror of https://github.com/mikegerber/ocrd_calamari.git synced 2025-06-09 19:59:53 +02:00
No description
Find a file
2020-02-03 15:33:11 +01:00
.circleci 🐛 CircleCI: Try upgrading pip 2020-01-31 10:46:50 +01:00
.idea 🔧 Add PyCharm project files 2019-12-05 12:33:33 +01:00
ocrd_calamari 📝 Document why we are using Unicode text segmentation to produce word results 2020-02-03 15:33:11 +01:00
test Remove broken __main__ handling (stick to pytest) 2019-12-05 13:47:31 +01:00
.coveragerc Only do the coverage on our code 2019-12-03 17:11:09 +01:00
.gitignore Fix tests by 1. binarizing and 2. use the GT4HistOCR model 2019-12-05 13:02:20 +01:00
Dockerfile Dockerfile 2019-10-26 22:12:16 +02:00
LICENSE Initial commit 2018-07-23 15:34:19 +02:00
Makefile circle: set locale to a UTF-8 variant so python doesn't fall back to ascii 2020-01-15 18:12:36 +01:00
ocrd-tool.json . 2018-07-26 19:09:07 +02:00
README-DEV.md 📝 README-DEV: Document installing test requirements 2020-01-30 19:06:42 +01:00
README.md 📝 README: Add information about the new glyph and word segmentation 2020-02-03 15:32:26 +01:00
requirements-test.txt Use GT segmentation to test 2019-12-03 17:11:09 +01:00
requirements.txt Include proper word + glyph segmentation 2020-02-03 12:22:01 +01:00
setup.py 📦 v0.0.3 – To fix version inconsistency 2019-12-02 18:23:00 +01:00

ocrd_calamari

Recognize text using Calamari OCR.

image image image

Introduction

This offers a OCR-D compliant workspace processor for some of the functionality of Calamari OCR.

This processor only operates on the text line level and so needs a line segmentation (and by extension a binarized image) as its input.

In addition to the line text it also outputs glyph segmentation including per-glyph confidence values and per-glyph alternative predictions as provided by the Calamari OCR engine. Note that while Calamari does not provide word segmentation, this processor produces word segmentation inferred from Unicode text segmentation and the glyph positions. The provided glyph and word segmentation can be used for text extraction and highlighting, but is probably not useful for further image-based processing.

Installation

From PyPI

pip install ocrd_calamari

From Repo

pip install .

Install models

Download models trained on GT4HistOCR data:

make gt4histocr-calamari
ls gt4histocr-calamari

Example Usage

ocrd-calamari-recognize -p test-parameters.json -m mets.xml -I OCR-D-SEG-LINE -O OCR-D-OCR-CALAMARI

With test-parameters.json:

{
    "checkpoint": "/path/to/some/trained/models/*.ckpt.json"
}

Development & Testing

For information regarding development and testing, please see README-DEV.md.