|
|
|
@ -59,18 +59,14 @@ do_binarization() {
|
|
|
|
|
ocrd_olena_binarize_parameters='{
|
|
|
|
|
"impl": "sauvola-ms-split"
|
|
|
|
|
}'
|
|
|
|
|
ocrd workspace remove-group -rf OCR-D-IMG-BINPAGE
|
|
|
|
|
ocrd workspace remove-group -rf OCR-D-IMG-BIN
|
|
|
|
|
ocrd-olena-binarize -I $INPUT_FILE_GRP -O OCR-D-IMG-BINPAGE,OCR-D-IMG-BIN -p "$ocrd_olena_binarize_parameters"
|
|
|
|
|
ocrd-olena-binarize --overwrite -I $INPUT_FILE_GRP -O OCR-D-IMG-BINPAGE,OCR-D-IMG-BIN -p "$ocrd_olena_binarize_parameters"
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
do_linesegmentation_tesserocr() {
|
|
|
|
|
# Segment the lines in the binarized images
|
|
|
|
|
|
|
|
|
|
ocrd workspace remove-group -rf OCR-D-SEG-REGION
|
|
|
|
|
ocrd workspace remove-group -rf OCR-D-SEG-LINE
|
|
|
|
|
ocrd-tesserocr-segment-region -I OCR-D-IMG-BINPAGE -O OCR-D-SEG-REGION
|
|
|
|
|
ocrd-tesserocr-segment-line -I OCR-D-SEG-REGION -O OCR-D-SEG-LINE
|
|
|
|
|
ocrd-tesserocr-segment-region --overwrite -I OCR-D-IMG-BINPAGE -O OCR-D-SEG-REGION
|
|
|
|
|
ocrd-tesserocr-segment-line --overwrite -I OCR-D-SEG-REGION -O OCR-D-SEG-LINE
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
do_linesegmentation_sbb() {
|
|
|
|
@ -79,9 +75,7 @@ do_linesegmentation_sbb() {
|
|
|
|
|
ocrd_sbb_textline_detector_parameters='{
|
|
|
|
|
"model": "/var/lib/textline_detection"
|
|
|
|
|
}'
|
|
|
|
|
ocrd workspace remove-group -rf OCR-D-SEG-REGION
|
|
|
|
|
ocrd workspace remove-group -rf OCR-D-SEG-LINE
|
|
|
|
|
ocrd-sbb-textline-detector -I OCR-D-IMG-BINPAGE -O OCR-D-SEG-LINE -p "$ocrd_sbb_textline_detector_parameters"
|
|
|
|
|
ocrd-sbb-textline-detector --overwrite -I OCR-D-IMG-BINPAGE -O OCR-D-SEG-LINE -p "$ocrd_sbb_textline_detector_parameters"
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
do_ocr() {
|
|
|
|
@ -91,8 +85,7 @@ do_ocr() {
|
|
|
|
|
"model": "GT4HistOCR_2000000",
|
|
|
|
|
"textequiv_level": "'$TEXTEQUIV_LEVEL'"
|
|
|
|
|
}'
|
|
|
|
|
ocrd workspace remove-group -rf OCR-D-OCR-TESS
|
|
|
|
|
ocrd-tesserocr-recognize -I OCR-D-SEG-LINE -O OCR-D-OCR-TESS -p "$ocrd_tesserocr_recognize_parameters"
|
|
|
|
|
ocrd-tesserocr-recognize --overwrite -I OCR-D-SEG-LINE -O OCR-D-OCR-TESS -p "$ocrd_tesserocr_recognize_parameters"
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
do_ocr_calamari() {
|
|
|
|
@ -102,8 +95,7 @@ do_ocr_calamari() {
|
|
|
|
|
"checkpoint": "/var/lib/calamari-models/GT4HistOCR/2019-07-22T15:49+0200/*.ckpt.json",
|
|
|
|
|
"textequiv_level": "'$TEXTEQUIV_LEVEL'"
|
|
|
|
|
}'
|
|
|
|
|
ocrd workspace remove-group -rf OCR-D-OCR-CALAMARI
|
|
|
|
|
ocrd-calamari-recognize -I OCR-D-SEG-LINE -O OCR-D-OCR-CALAMARI -p "$ocrd_calamari_recognize_parameters"
|
|
|
|
|
ocrd-calamari-recognize --overwrite -I OCR-D-SEG-LINE -O OCR-D-OCR-CALAMARI -p "$ocrd_calamari_recognize_parameters"
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
page_validate_xml() {
|
|
|
|
@ -159,8 +151,7 @@ main() {
|
|
|
|
|
do_validate
|
|
|
|
|
|
|
|
|
|
if ocrd workspace list-group | grep -q OCR-D-GT-PAGE; then
|
|
|
|
|
ocrd workspace remove-group -rf $ocr_filegrp-EVAL
|
|
|
|
|
ocrd-dinglehopper -I OCR-D-GT-PAGE,$ocr_filegrp -O $ocr_filegrp-EVAL
|
|
|
|
|
ocrd-dinglehopper --overwrite -I OCR-D-GT-PAGE,$ocr_filegrp -O $ocr_filegrp-EVAL
|
|
|
|
|
fi
|
|
|
|
|
|
|
|
|
|
done
|
|
|
|
|