sporada/ocr
1
0
Fork 0
mirror of https://gitlab.ub.uni-bielefeld.de/sfb1288inf/ocr.git synced 2024-12-26 19:14:18 +00:00
Code Issues Projects Releases Wiki Activity

Add back german language models

Browse Source
This commit is contained in:
Patrick Jentsch 2021-03-17 14:26:24 +01:00
parent 947658a7d8
commit 6db7f70446
2 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
Dockerfile
View File

@ -68,7 +68,7 @@ RUN wget --no-check-certificate --quiet \
&& cd - > /dev/null \ && cd - > /dev/null \
&& rm -r "tesseract-${TESSERACT_VERSION}" "${TESSERACT_VERSION}.tar.gz" && rm -r "tesseract-${TESSERACT_VERSION}" "${TESSERACT_VERSION}.tar.gz"
ENV TESSERACT_MODELS="ara,chi_tra,dan,ell,eng,enm,fra,frk,frm,ita,por,rus,spa" ENV TESSERACT_MODELS="ara,chi_tra,dan,deu,ell,eng,enm,fra,frk,frm,ita,por,rus,spa"
ENV TESSDATA_BEST_VERSION=4.1.0 ENV TESSDATA_BEST_VERSION=4.1.0
RUN wget --no-check-certificate --quiet \ RUN wget --no-check-certificate --quiet \
"https://github.com/tesseract-ocr/tessdata_best/archive/${TESSDATA_BEST_VERSION}.tar.gz" \ "https://github.com/tesseract-ocr/tessdata_best/archive/${TESSDATA_BEST_VERSION}.tar.gz" \

2
ocr
View File

@ -322,7 +322,7 @@ def parse_args():
type=int) type=int)
parser.add_argument('--n-cores', parser.add_argument('--n-cores',
default=min(4, multiprocessing.cpu_count()), default=min(4, multiprocessing.cpu_count()),
help='Number of CPU threads to be used', # noqa help='Number of CPU threads to be used',
type=int) type=int)
parser.add_argument('--zip', parser.add_argument('--zip',
help='Create one zip file per filetype') help='Create one zip file per filetype')