🚧 Add preliminary support for ocrd_calamari's Calamari 1 version

This commit is contained in:
Gerber, Mike 2020-11-17 10:00:38 +01:00
parent f2ed0b5cbf
commit b6bbc7ca3a
3 changed files with 34 additions and 2 deletions

6
build
View file

@ -7,6 +7,7 @@ self_dir=`dirname "$self"`
DATA_SUBDIR=data
get_from_annex() {
annex_get 'calamari-models/GT4HistOCR/2019-07-22T15_49+0200/*.ckpt*'
annex_get 'calamari-models/GT4HistOCR/2019-12-11T11_10+0100/*.ckpt*'
annex_get 'tesseract-models/GT4HistOCR/*.traineddata'
annex_get 'textline_detection/*.h5'
annex_get 'mirror/github.com/tesseract-ocr/tessdata_best/archive/4.0.0-repacked.tar.gz'
@ -14,6 +15,7 @@ get_from_annex() {
}
get_from_web() {
download_to 'https://qurator-data.de/calamari-models/GT4HistOCR/model.tar.xz' 'calamari-models/GT4HistOCR/2019-07-22T15_49+0200'
# XXX Calamari 1 model
download_to 'https://qurator-data.de/tesseract-models/GT4HistOCR/models.tar' 'tesseract-models/GT4HistOCR'
download_to 'https://qurator-data.de/sbb_textline_detector/models.tar.gz' 'textline_detection'
download_to --strip-components 1 'https://qurator-data.de/sbb_binarization/models.tar.gz' 'sbb_binarization'
@ -32,3 +34,7 @@ docker build --cache-from=my_ocrd_workflow-sbb_textline_detector -t my_ocrd_work
docker build --cache-from=my_ocrd_workflow-sbb_binarization -t my_ocrd_workflow-sbb_binarization -f Dockerfile-sbb_binarization .
docker build --cache-from=my_ocrd_workflow-ocrd_cis -t my_ocrd_workflow-ocrd_cis -f Dockerfile-ocrd_cis .
docker build --cache-from=my_ocrd_workflow-ocrd_fileformat -t my_ocrd_workflow-ocrd_fileformat -f Dockerfile-ocrd_fileformat .
# XXX
docker build --cache-from=my_ocrd_workflow-ocrd_calamari-feat-update-calamari1 -t my_ocrd_workflow-ocrd_calamari-feat-update-calamari1 -f Dockerfile-ocrd_calamari-feat-update-calamari1 .