TheAnachronism/docspell
1
0
Fork 0
mirror of https://github.com/TheAnachronism/docspell.git synced 2025-11-03 18:00:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

Use different japanese train files for tesseract

Browse Source 
They seem to work better as suggested here:
https://github.com/tesseract-ocr/tessdata/issues/119

Refs: #973
This commit is contained in:
eikek
2021-08-13 16:44:56 +02:00
parent f79aa447b5
commit 326cf1c087
1 changed files with 6 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
docker/dockerfiles/joex.dockerfile
View File

@@ -63,6 +63,12 @@ RUN wget ${joex_url:-https://github.com/eikek/docspell/releases/download/v$versi
rm docspell-joex-*.zip && \
ln -snf docspell-joex-* docspell-joex
# Using these data files for japanese, because they work better. See #973
RUN \
wget https://raw.githubusercontent.com/tesseract-ocr/tessdata_fast/master/jpn_vert.traineddata && \
wget https://raw.githubusercontent.com/tesseract-ocr/tessdata_fast/master/jpn.traineddata && \
mv jpn*.traineddata /usr/share/tessdata
COPY joex-entrypoint.sh /opt/joex-entrypoint.sh
ENTRYPOINT ["/opt/joex-entrypoint.sh", "-J-XX:+UseG1GC"]