You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Go to file
Gerber, Mike c7ad6eb724 📦 v1.0.5 2 years ago
.circleci ✔️ CircleCI: Use a test matrix to test on different Python versions 2 years ago
.idea 🔧 Add PyCharm project files 5 years ago
ocrd_calamari 📦 v1.0.5 2 years ago
test 🔐 Use a safe temporary directory for the test workspace 2 years ago
.coveragerc Only do the coverage on our code 5 years ago
.gitignore fix typos 3 years ago
Dockerfile 🐳 Docker: Run pip3 check for good measure 4 years ago
LICENSE 📄 Update license (Fixes #35) 4 years ago
Makefile ✔️ Move installing deps to Makefile 2 years ago 📝 README-DEV: Update wrt ocrd-tool.json and CI 2 years ago 🧹 README: Do not mention obsolete + unsupported checkpoint parameter 2 years ago
ocrd-tool.json . 6 years ago
requirements-test.txt Use GT segmentation to test 5 years ago
requirements.txt 🐛 Fix TensorFlow (and h3py) dependency 3 years ago read version from ocrd-tool.json 3 years ago


Recognize text using Calamari OCR.

image image image


ocrd_calamari offers a OCR-D compliant workspace processor for the functionality of Calamari OCR. It uses OCR-D workspaces (METS) with PAGE XML documents as input and output.

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 may also output word and glyph segmentation including per-glyph confidence values and per-glyph alternative predictions as provided by the Calamari OCR engine, using a textequiv_level of word or glyph. Note that while Calamari does not provide word segmentation, this processor produces word segmentation inferred from 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.

Example output as viewed in PAGE Viewer


From PyPI

pip install ocrd_calamari

From Repo

pip install .

Install models

Download models trained on GT4HistOCR data:

make gt4histocr-calamari1
ls gt4histocr-calamari1

Manual download: model.tar.xz

Example Usage

Before using ocrd-calamari-recognize get some example data and model, and prepare the document for OCR:

# Download model and example data
make gt4histocr-calamari1
make actevedef_718448162

# Create binarized images and line segmentation using other OCR-D projects
cd actevedef_718448162
ocrd-olena-binarize -P impl sauvola-ms-split -I OCR-D-IMG -O OCR-D-IMG-BIN
ocrd-tesserocr-segment-region -I OCR-D-IMG-BIN -O OCR-D-SEG-REGION
ocrd-tesserocr-segment-line -I OCR-D-SEG-REGION -O OCR-D-SEG-LINE

Finally recognize the text using ocrd_calamari and the downloaded model:

ocrd-calamari-recognize -P checkpoint_dir "../gt4histocr-calamari1" -I OCR-D-SEG-LINE -O OCR-D-OCR-CALAMARI

You may want to have a look at the ocrd-tool.json descriptions for additional parameters and default values.

Development & Testing

For information regarding development and testing, please see