mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2024-12-05 02:47:00 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

Use different font list for grc training

Browse Source 
This font list contains a selection fonts produced by the Greek Font
Society <http://greekfontsociety.gr>, and is the result of testing
with a large corpus of a variety of scanned works.
This commit is contained in:
Nick White 2015-09-10 15:34:44 +01:00 committed by Zdenko Podobný
parent b848caa151
commit 890ec2876b
1 changed files with 29 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
training/language-specific.sh
View File

@ -353,6 +353,30 @@ GREEK_FONTS=( \
"VL PGothic" \
)
ANCIENT_GREEK_FONTS=( \
"GFS Artemisia \
"GFS Artemisia Bold" \
"GFS Artemisia Bold Italic" \
"GFS Artemisia Italic" \
"GFS Bodoni" \
"GFS Bodoni Bold" \
"GFS Bodoni Bold Italic" \
"GFS Bodoni Italic" \
"GFS Didot" \
"GFS Didot Bold" \
"GFS Didot Bold Italic" \
"GFS Didot Italic" \
"GFS DidotClassic" \
"GFS Neohellenic" \
"GFS Neohellenic Bold" \
"GFS Neohellenic Bold Italic" \
"GFS Neohellenic Italic" \
"GFS Philostratos" \
"GFS Porson" \
"GFS Pyrsos" \
"GFS Solomos" \
)
ARABIC_FONTS=( \
"Arabic Transparent Bold" \
"Arabic Transparent" \
@ -1039,10 +1063,14 @@ set_lang_specific_parameters() {
chr ) test -z "$FONTS" && FONTS=( "${NORTH_AMERICAN_ABORIGINAL_FONTS[@]}" \
"Noto Sans Cherokee" \
) ;;
ell | grc )
ell )
NUMBER_DAWG_FACTOR=0.05
WORD_DAWG_FACTOR=0.08
test -z "$FONTS" && FONTS=( "${GREEK_FONTS[@]}" ) ;;
grc )
NUMBER_DAWG_FACTOR=0.05
WORD_DAWG_FACTOR=0.08
test -z "$FONTS" && FONTS=( "${ANCIENT_GREEK_FONTS[@]}" ) ;;
hye ) test -z "$FONTS" && FONTS=( "${ARMENIAN_FONTS[@]}" ) ;;
iku ) test -z "$FONTS" && FONTS=( "${NORTH_AMERICAN_ABORIGINAL_FONTS[@]}" ) ;;
kat) test -z "$FONTS" && FONTS=( "${GEORGIAN_FONTS[@]}" ) ;;