This project creates a docker image for Kitodo Production 3.7.1 and attach OCR-D maximum image to it for the OCR step.
In this version, the provided Example of workflow is modified to support Operandi-Integration-Script: https://github.com/subugoe/Operandi-Integration-Script/
OLA-HD step is resposible of uploading the OCR-D results to OLA_HD Server: https://ola-hd.ocr-d.de/ OCR-D is supported locally and on HPC servers using Operandi project: https://github.com/subugoe/operandi
To use the OCR-D local processing -l flag must be added to Operandi(OCR) step in the workflow. To understand more, please read https://github.com/subugoe/Operandi-Integration-Script/
The project folder structure should contain at least following sub-folders:
, where images/master_(processtitle)_media is the main images directory and ocr/alto is the OCR-D results directory. Thus, the Operandi(OCR) step needs those 2 paths to work.
Also, any other file group sub-folder is optional and up to the user to decide.
- Set
OPERANDI_USER_PASS="operandi_username:operandi_password"andOLA_USR="ola_hd_username:ola_hd_password"values in the .env file - Then
./run.sh
Kitodo Production can be accessed at http://localhost:8080/kitodo with initial credentials testAdmin / test
This version of Kitodo has some rendering issues with some browsers, but works the best with Firefox browser