mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2024-12-18 11:28:51 +08:00
Code Issues Actions 10 Packages Projects Releases Wiki Activity

Use different font list and exposures for "lat" language training

Browse Source
This commit is contained in:
Ryan Baumann 2016-01-04 11:48:02 -05:00 committed by Zdenko Podobný
parent b60bb806bf
commit 52d9ed4d69
1 changed files with 32 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
training/language-specific.sh
View File

@ -69,6 +69,35 @@ LATIN_FONTS=(
"DejaVu Sans Ultra-Light" \
)
# List of fonts for printed/neo-Latin ('lat' language code, different from Latin script)
NEOLATIN_FONTS=(
"GFS Bodoni" \
"GFS Bodoni Bold" \
"GFS Bodoni Italic" \
"GFS Bodoni Bold Italic" \
"GFS Didot" \
"GFS Didot Bold" \
"GFS Didot Italic" \
"GFS Didot Bold Italic" \
"Cardo" \
"Cardo Bold" \
"Cardo Italic" \
"Wyld" \
"Wyld Italic" \
"EB Garamond" \
"EB Garamond Italic" \
"IM FELL DW Pica PRO" \
"IM FELL English PRO" \
"IM FELL Double Pica PRO" \
"IM FELL French Canon PRO" \
"IM FELL Great Primer PRO" \
"IM FELL DW Pica PRO Italic" \
"IM FELL English PRO Italic" \
"IM FELL Double Pica PRO Italic" \
"IM FELL French Canon PRO Italic" \
"IM FELL Great Primer PRO Italic" \
)
EARLY_LATIN_FONTS=(
"${FRAKTUR_FONTS[@]}" \
"${LATIN_FONTS[@]}" \
@ -853,6 +882,9 @@ set_lang_specific_parameters() {
FILTER_ARGUMENTS="--make_early_language_variant=ita"
TEXT2IMAGE_EXTRA_ARGS=" --ligatures" # Add ligatures when supported.
test -z "$FONTS" && FONTS=( "${EARLY_LATIN_FONTS[@]}" );;
lat )
test -z "$EXPOSURES" && EXPOSURES="-3 -2 -1 0 1 2 3"
test -z "$FONTS" && FONTS=( "${NEOLATIN_FONTS[@]}" ) ;;
spa_old )
TEXT_CORPUS="${FLAGS_webtext_prefix}/spa.corpus.txt"
# Make long-s substitutions for Early Spanish text
@ -893,7 +925,6 @@ set_lang_specific_parameters() {
isl ) ;;
ita ) ;;
jav ) ;;
lat ) ;;
lav ) ;;
lit ) ;;
mlt ) ;;