Setup guide for pilot libraries

OCR-D’s software is a modular collection of many projects (called modules) with many tools per module (called processors) that you can combine freely to achieve the workflow best suited for OCRing your content.

All OCR-D modules follow the same interface and common design patterns. So once you understand how to install and use one project, you know how to install and use all of them.

Installation

There are three ways to install OCR-D modules:

  1. Using the ocrd/all Docker module collection (recommended)
  2. Using ocrd/all to install OCR-D modules locally
  3. Installing modules indivudally via Docker or natively (not recommended)

We recommend using the Docker image since this does not require any changes to the host system besides installing Docker.

We do not recommend installing modules individually because it can be difficult to keep the software up-to-date and ensure that they are at working and interoperable versions.

ocrd_all

The ocrd_all project is an effort by the OCR-D community, now maintained by the OCR-D coordination team. It streamlines the native installation of OCR-D modules with a versatile Makefile approach. Besides allowing local installation of the full OCR-D stack, it is also the base for the ocrd/all Docker images available from DockerHub that contain the full stack of OCR-D modules ready for deployment.

Technically, ocrd_all is a Git repository that keeps all the necessary software as Git submodules at specific revisions. This way, the software tools are known to be at a stable version and guaranteed to be interoperable with one another.

ocrd_all via Docker

mini medi maxi

There are three versions of the ocrd/all image: minimum, medium and maximum. They differ in which modules are included and hence the size of the image. Only use the minimum or medium images if you are certain that you do not need the full OCR-D stack for your workflows, otherwise we encourage you to use the large but complete maximum image.

Check this table to see which modules are included in which version:

Module minimum medium maximum
core
ocrd_cis
ocrd_im6convert
ocrd_repair_inconsistencies
ocrd_tesserocr
tesserocr
workflow-configuration
cor-asv-ann -
dinglehopper -
format-converters -
ocrd_calamari -
ocrd_keraslm -
ocrd_olena -
ocrd_segment -
tesseract -
ocrd_anybaseocr - -
ocrd_kraken - -
ocrd_ocropy - -
ocrd_pc_segmentation - -
ocrd_typegroups_classifier - -
sbb_textline_detector - -
cor-asv-fst - -

Fetch docker image

To fetch the maximum version of the ocrd/all Docker image:

docker pull ocrd/all:maximum

Replace maximum accordingly if you want the minimum or medium variant.

If no specific version is chosen, latest is selected by default, which is equivalent to medium.

Updating docker image

To update the docker images to their latest version, just run the docker pull command again:

docker pull ocrd/all:<version>

This can even be set up as a cron-job to ensure the image is always up-to-date.

Translating native commands to docker calls

In the documentation, both of the OCR-D coordination project as well as the documentation of the individual OCR-D modules, you will find native commands, i.e. command line calls that expect the software to be installed natively. These are simple to translate to commands based on the docker images by prepending the boilerplate telling Docker which image to use, which user to run as, which files to bind to a container path etc.

For example a call to ocrd-tesserocr-binarize might natively look like this:

ocrd-tesserocr-segment-region -I OCR-D-IMG -O OCR-D-SEG-BLOCK

To run it with the [ocrd/all:maximum] Docker container:

docker run -u $(id -u) -v $PWD:/data -w /data -- ocrd/all:maximum ocrd-tesserocr-segment-region -I OCR-D-IMG -O OCR-D-SEG-BLOCK
           \_________/ \___________/ \______/ \_________________/ \___________________________________________________________/
              (1)          (2)         (3)          (4)                            (5)

It can also be useful to delete the container after creation with the --rm parameter.

ocrd_all natively

The ocrd_all project contains a sophisticated Makefile to install or compile prerequisites as necessary, set up a virtualenv, install the core software, install OCR-D modules and more. Detailed documentation can be found in its README.

Prerequisites

There are some system requirements for ocrd_all.

You need to have git and make installed to make use of ocrd_all:

sudo apt install git make

It is easiest to install all the possible system requirements by calling make deps-ubuntu as root:

sudo make deps-ubuntu

Cloning the repository

Clone the repository and all its submodules:

git clone --recursive https://github.com/OCR-D/ocrd_all
cd ocrd_all

Installing with ocrd_all

You can either install

  1. all the software at once with the all target (equivalent to the maximum Docker version)
  2. modules individually by using an executable from that module as the target or :
  3. modules invidually by using the project name for the OCRD_MODULES variable:
make all                       # Installs all the software (recommended)

make ocrd-tesserocr-binarize   # Install ocrd_tesserocr which contains ocrd-tesserocr-binarize
make ocrd-cis-ocropy-binarize  # Install ocrd_cis  which contains ocrd-cis-ocropy-binarize

make all OCRD_MODULES="ocrd_tesserocr ocrd_cis"  # Will install both ocrd_tesserocr and ocrd_cis

Updating the repository

As ocrd_all is in active development, it is wise to regularly update the repository and its submodules:

git pull
make modules

Individual installation

With all variants of individual module installation, it will be up to you to keep the repositories up-to-date and installed. We therefore discourage individual installation of modules and recommend using ocrd_all as outlined above..

Individual Docker container

This is the best option if you want full control over which modules you actually intend to use while still profiting from the simple installation of Docker containers.

You need to have Docker

All OCR-D modules are also published as Docker containers to DockerHub. To find the docker image for a module, replace the ocrd_ prefix with ocrd/:

docker pull ocrd/tesserocr  # Installs ocrd_tesserocr
docker pull ocrd/olena  # Installs ocrd_olena

To run the containers, please see the notes on translating native command line calls to docker calls above. Make sure the image name matches the executable. For example to run the same example in the dedicated ocrd_tesserocr container:

docker run -u $(id -u) -w /data -v $PWD:/data -- ocrd/tesserocr ocrd-tesserocr-segment-region -I OCR-D-IMG -O OCR-D-SEG-BLOCK-DOCKER

Native installation

NOTE

ocrd_tesserocr requires tesseract-ocr >= 4.1.0. But the Tesseract packages bundled with Ubuntu < 19.10 are too old. If you are on Ubuntu 18.04 LTS, please enable Alexander Pozdnyakov PPA, which has up-to-date builds of tesseract and its dependecies:

sudo add-apt-repository ppa:alex-p/tesseract-ocr
sudo apt-get update

virtualenv

First install Python 3 and venv:

sudo apt install python3 python3-venv
# If you haven't created the venv yet:
python3 -m venv ~/venv
# Activate the venv
source ~/venv/bin/activate

Once you have activated the virtualenv, you should see (venv) prepended to your shell prompt.

From PyPI

This is the best option if you want to use the stable, released version of individual modules.

However, many modules require a number of non-Python (system) packages. For the exact list of packages you need to look at the README of the module in question. (If you are not on Ubuntu >= 18.04, then your requirements may deviate from that.)

For example to install ocrd_tesserocr from PyPI:

sudo apt-get install git python3 python3-pip python3-venv libtesseract-dev libleptonica-dev tesseract-ocr-eng tesseract-ocr wget
pip3 install ocrd_tesserocr

Many ocrd modules conventionally contain a Makefile with a deps-ubuntu target that can handle calls to apt-get for you:

sudo make deps-ubuntu

From git

This is the best option if you want to change the source code or install the latest, unpublished changes.

git clone https://github.com/OCR-D/ocrd_tesserocr
cd ocrd_tesserocr
sudo make deps-ubuntu # or manually with apt-get
make deps             # or pip3 install -r requirements
make install          # or pip3 install .

If you intend to develop a module, it is best to install the module editable:

pip install -e .

This way, you won’t have to reinstall after making changes.

Testing the installation

For example, let’s fetch a document from the OCR-D GT Repo:

wget 'https://ocr-d-repo.scc.kit.edu/api/v1/dataresources/736a2f9a-92c6-4fe3-a457-edfa3eab1fe3/data/wundt_grundriss_1896.ocrd.zip'
unzip wundt_grundriss_1896.ocrd.zip
cd data

Test native installation

This section applies if you installed the software natively, either via ocrd_all or on a per-module basis.

First, activate your venv:

# Activate the venv
source ~/venv/bin/activate

Let’s segment the images in file group OCR-D-IMG into regions (creating a first PAGE-XML file group OCR-D-SEG-BLOCK):

ocrd-tesserocr-segment-region -I OCR-D-IMG -O OCR-D-SEG-BLOCK

Test Docker installation

This section applies if you installed the software as docker container(s), either via ocrd_all or on a per-module basis.

You can spin up a docker container, mounting the current working directory like this:

docker run -u $(id -u) -w /data -v $PWD:/data -- ocrd/all:maximum ocrd-tesserocr-segment-region -I OCR-D-IMG -O OCR-D-SEG-BLOCK-DOCKER

Note that the CLI is exactly the same, the only difference is the prefix to instruct Docker, as explained above

Running a small workflow

With PyPI and workflow engine from core

The core package will be installed with every OCR-D module package, but you can also install it manually:

pip3 install ocrd

Its CLI ocrd contains a simple workflow engine, available with the ocrd process command, which allows you to chain multiple OCR-D processor calls into simple sequential workflows.

For example, let’s combine the ocropy-based binarization of the ocrd_cis module project with the segmentation and recognition in ocrd_tesserocr.

First, install ocrd_cis, too:

# Install ocrd_cis
pip3 install ocrd_cis # with pip

Next, install a suitable OCR model for Tesseract:

# Install OCR model into Tesseract datapath
sudo apt-get install tesseract-ocr-script-frak

Now we can define the workflow (as a list of processor calls in abbreviated form, and a number of parameter files where defaults are insufficient):

# Create parameter files
echo '{ "model": "Fraktur" }' > param-tess-fraktur.json

# Run workflow
ocrd process \
  'cis-ocropy-binarize -I OCR-D-IMG -O OCR-D-SEG-PAGE' \
  'tesserocr-segment-region -I OCR-D-SEG-PAGE -O OCR-D-SEG-BLOCK' \
  'tesserocr-segment-line -I OCR-D-SEG-BLOCK -O OCR-D-SEG-LINE' \
  'tesserocr-recognize -I OCR-D-SEG-LINE -O OCR-D-OCR-TESSEROCR -p param-tess-fraktur.json'