Module 0 Software needed for the OCR workshop

Introduction

This is a list of software requirements to successfully complete the practice sessions and maximize your learning experience of the workshop on
OCR and postcorrection of early printings for digital humanities
at LMU, Munich 2015-09-14/15

on OS: Linux is your best option; MacOS can be used almost as well (some problems might arise with OCRopus installation); Windows users will not be able to train or run OCRopus models (they may consider a virtualbox Linux installation)
our focus is on open source software
for Linux, almost all software is available in its package repositories and can be installed from there
you will be optimally prepared if the software is running on your laptop before the workshop begins, and you know how to use it
the software packages mentioned below are part of a complete OCR toolchain - if you miss some parts because of installation problems, we will provide you with suitable input data for each separate step

we will treat Tesseract and OCRopus (open source engines) as well as Abbyy Finereader (commercial)
Tesseract and OCRopus can be downloaded and installed locally; Abbyy will provide a demo key for its online service
both Tesseract and OCRopus have recently been moved from Google Code to GitHub (click on blue inline links and follow the installation procedures):
- Tesseract: available for Linux, Mac, Windows
- OCRopus: Linux (Mac)
- OCRopus (now called Ocropy) can be installed in your home directory:
```
    python setup.py install --home=~/<install-dir>
    export PATH=$HOME/<install-dir>/bin:$PATH
    export PYTHONPATH=$HOME/<install-dir>/lib/python
```
if you want to install OCRopus in a docker environment (Mac):
- use the Ocrocis wrapper
don't forget to also download some Tesseract training files for languages, called lang.traineddata:
- deu_frak: German Fraktur
- grc: Ancient Greek from Nick White
- lat: Latin from Ryan Baumann

download the binary distribution of PoCoTo
this will download a zipped archive file ocrcorrection.zip.
extract (unzip) this archive to a convenient place somewhere in your user directory
this will create a folder ocrcorrection
in the folder ocrcorrection/bin, identify the appropriate executable for your operating system:
- MS Windows: either ocrcorrection.exe or ocrcorrection64.exe
- otherwise it is the file ocrcorrection
before you start the application, make sure that the Java Runtime Environment (jre) is installed on your system
PoCoTO is described in detail in its manual which may be consulted for any questions

split pdf into single page images:
- pdftk from PDFtk
further pdf tools:
- pdftoppm, pdfimages from Xpdf tools, or (Linux) from poppler-utils package
format conversion:
- convert from ImageMagick
further preprocessing: ScanTailord
learn how to use ScanTailor

we need to be able to evaluate OCR output against ground truth
a widely used tool collection is the Rice/Nartker UNLV/ISRI OCR evaluation toolkit
user guide and source code with UTF-8 wrapper from Nick White
needs to be compiled locally