Run Calamari OCR

This commit is contained in:
Gerber, Mike 2019-08-21 11:54:01 +02:00
parent 001e62f54a
commit 0bc06c2fad
7 changed files with 29 additions and 1 deletions

View file

@ -81,6 +81,14 @@ do_ocr() {
-p <(echo $ocrd_tesserocr_recognize_parameters)
}
do_ocr_calamari() {
ocrd_calamari_recognize_parameters='{ "checkpoint": "/var/lib/calamari-models/GT4HistOCR/*.ckpt.json" }'
remove_filegrp OCR-D-OCR-CALAMARI mets.xml
ocrd-calamari-recognize -l $LOG_LEVEL \
-m mets.xml -I OCR-D-SEG-LINE -O OCR-D-OCR-CALAMARI \
-p <(echo $ocrd_calamari_recognize_parameters)
}
page_validate_xml() {
# Validate all PAGE XML against the XML schema
@ -141,6 +149,8 @@ page_validate_xml OCR-D-SEG-REGION
page_validate_xml OCR-D-SEG-LINE
do_validate
do_ocr_calamari
do_ocr
page_validate_xml OCR-D-OCR-TESS
page_workaround_remove_conf OCR-D-OCR-TESS