sk-spell

podpora slovenčiny v Open Source programoch

Update of language files for tesseract-ocr 3.04   

back to tesseract-ocr-en

There was huge update of tesseract-ocr language files on 24.06.2015 – 98 traineddata were updated or first uploaded. At the moment 105 of language or language version are supported (+2 special modules osd and equ). The corresponding source training data where commited into langdata repository.

ara, eng, hin, kor, osd, equ traineddata are NOT updated due to regression. The other regressions are mostly fixed, with some dramatic improvements particularly for Indic (like 20% for kan for example).

There was no update for cube files, because cube is dead end and will be removed after new classifier implementation.

language files

Language files are located in separate repository tessdata on github.com. Total file size is 1.2 GiB at the moment. If you plan to clone whole repository you need to calculate much more space because git keep all history (e.g. previous version of files) in local copy…

Therefore it would be more efficient to download only needed language files. I created small program get_tessdata.cpp that download and install desired file for you. It is not smart (e.g. there could be problem with proxy, or you can not choose file version), but I hope it can help. Usage is simple (after compilation ;-) ) – e.g.:

sudo ./get_tessdata -f fra.traineddata

Feel free to improve it/replace it ;-) and share it tesseract-ocr user forum.

Here is the list of available files at tessdata repository as of 29.06.2015:

lang code	lang name	file size	link
afr	Afrikaans	5.0 MiB	afr.traineddata
amh	Amharic	2.8 MiB	amh.traineddata
ara	Arabic	99.5 KiB	ara.cube.bigrams
ara	Arabic	4.0 B	ara.cube.fold
ara	Arabic	241.0 B	ara.cube.lm
ara	Arabic	820.7 KiB	ara.cube.nn
ara	Arabic	251.0 B	ara.cube.params
ara	Arabic	19.1 MiB	ara.cube.size
ara	Arabic	1.2 MiB	ara.cube.word-freq
ara	Arabic	6.0 MiB	ara.traineddata
asm	Assamese	15.1 MiB	asm.traineddata
aze	Azerbaijani	6.3 MiB	aze.traineddata
aze_cyrl	Azerbaijani – Cyrilic	2.7 MiB	aze_cyrl.traineddata
bel	Belarusian	6.5 MiB	bel.traineddata
ben	Bengali	14.8 MiB	ben.traineddata
bod	Tibetan	24.1 MiB	bod.traineddata
bos	Bosnian	5.2 MiB	bos.traineddata
bul	Bulgarian	5.7 MiB	bul.traineddata
cat	Catalan; Valencian	5.1 MiB	cat.traineddata
ceb	Cebuano	1.6 MiB	ceb.traineddata
ces	Czech	11.3 MiB	ces.traineddata
chi_sim	Chinese – Simplified	40.1 MiB	chi_sim.traineddata
chi_tra	Chinese – Traditional	54.1 MiB	chi_tra.traineddata
chr	Cherokee	1.0 MiB	chr.traineddata
cym	Welsh	3.6 MiB	cym.traineddata
dan	Danish	7.0 MiB	dan.traineddata
dan_frak	Danish – Fraktur	1.5 MiB	dan_frak.traineddata
deu	German	12.7 MiB	deu.traineddata
deu_frak	German – Fraktur	1.9 MiB	deu_frak.traineddata
dzo	Dzongkha	3.2 MiB	dzo.traineddata
ell	Greek, Modern (1453-)	5.2 MiB	ell.traineddata
eng	English	167.9 KiB	eng.cube.bigrams
eng	English	38.0 B	eng.cube.fold
eng	English	181.0 B	eng.cube.lm
eng	English	837.2 KiB	eng.cube.nn
eng	English	254.0 B	eng.cube.params
eng	English	12.4 MiB	eng.cube.size
eng	English	2.3 MiB	eng.cube.word-freq
eng	English	996.0 B	eng.tesseract_cube.nn
eng	English	20.9 MiB	eng.traineddata
enm	English, Middle (1100-1500)	2.0 MiB	enm.traineddata
epo	Esperanto	6.3 MiB	epo.traineddata
equ	Math / equation detection module	2.1 MiB	equ.traineddata
est	Estonian	9.2 MiB	est.traineddata
eus	Basque	4.7 MiB	eus.traineddata
fas	Persian	4.6 MiB	fas.traineddata
fin	Finnish	12.7 MiB	fin.traineddata
fra	French	127.0 KiB	fra.cube.bigrams
fra	French	59.0 B	fra.cube.fold
fra	French	301.0 B	fra.cube.lm
fra	French	949.5 KiB	fra.cube.nn
fra	French	242.0 B	fra.cube.params
fra	French	18.4 MiB	fra.cube.size
fra	French	2.8 MiB	fra.cube.word-freq
fra	French	660.0 B	fra.tesseract_cube.nn
fra	French	13.4 MiB	fra.traineddata
frk	Frankish	15.7 MiB	frk.traineddata
frm	French, Middle (ca.1400-1600)	15.1 MiB	frm.traineddata
gle	Irish	3.3 MiB	gle.traineddata
glg	Galician	5.3 MiB	glg.traineddata
grc	Greek, Ancient (to 1453)	4.9 MiB	grc.traineddata
guj	Gujarati	10.1 MiB	guj.traineddata
hat	Haitian; Haitian Creole	1.3 MiB	hat.traineddata
heb	Hebrew	4.1 MiB	heb.traineddata
hin	Hindi	67.4 KiB	hin.cube.bigrams
hin	Hindi	1.0 B	hin.cube.fold
hin	Hindi	211.0 B	hin.cube.lm
hin	Hindi	6.9 MiB	hin.cube.nn
hin	Hindi	262.0 B	hin.cube.params
hin	Hindi	1.2 MiB	hin.cube.word-freq
hin	Hindi	660.0 B	hin.tesseract_cube.nn
hin	Hindi	13.5 MiB	hin.traineddata
hrv	Croatian	8.7 MiB	hrv.traineddata
hun	Hungarian	11.6 MiB	hun.traineddata
iku	Inuktitut	971.9 KiB	iku.traineddata
ind	Indonesian	6.2 MiB	ind.traineddata
isl	Icelandic	5.8 MiB	isl.traineddata
ita	Italian	119.8 KiB	ita.cube.bigrams
ita	Italian	51.0 B	ita.cube.fold
ita	Italian	257.0 B	ita.cube.lm
ita	Italian	872.1 KiB	ita.cube.nn
ita	Italian	314.0 B	ita.cube.params
ita	Italian	13.3 MiB	ita.cube.size
ita	Italian	3.4 MiB	ita.cube.word-freq
ita	Italian	660.0 B	ita.tesseract_cube.nn
ita	Italian	13.6 MiB	ita.traineddata
ita_old	Italian – Old	13.4 MiB	ita_old.traineddata
jav	Javanese	4.2 MiB	jav.traineddata
jpn	Japanese	31.5 MiB	jpn.traineddata
kan	Kannada	34.0 MiB	kan.traineddata
kat	Georgian	5.9 MiB	kat.traineddata
kat_old	Georgian – Old	643.9 KiB	kat_old.traineddata
kaz	Kazakh	4.3 MiB	kaz.traineddata
khm	Central Khmer	46.6 MiB	khm.traineddata
kir	Kirghiz; Kyrgyz	5.2 MiB	kir.traineddata
kor	Korean	12.7 MiB	kor.traineddata
kur	Kurdish	1.9 MiB	kur.traineddata
lao	Lao	20.1 MiB	lao.traineddata
lat	Latin	5.7 MiB	lat.traineddata
lav	Latvian	7.4 MiB	lav.traineddata
lit	Lithuanian	8.5 MiB	lit.traineddata
mal	Malayalam	8.4 MiB	mal.traineddata
mar	Marathi	13.6 MiB	mar.traineddata
mkd	Macedonian	3.7 MiB	mkd.traineddata
mlt	Maltese	4.9 MiB	mlt.traineddata
msa	Malay	6.2 MiB	msa.traineddata
mya	Burmese	66.5 MiB	mya.traineddata
nep	Nepali	15.1 MiB	nep.traineddata
nld	Dutch; Flemish	16.3 MiB	nld.traineddata
nor	Norwegian	7.9 MiB	nor.traineddata
ori	Oriya	7.5 MiB	ori.traineddata
osd	Orientation and script detection module	10.1 MiB	osd.traineddata
pan	Panjabi; Punjabi	9.7 MiB	pan.traineddata
pol	Polish	13.3 MiB	pol.traineddata
por	Portuguese	12.3 MiB	por.traineddata
pus	Pushto; Pashto	2.4 MiB	pus.traineddata
ron	Romanian; Moldavian; Moldovan	7.6 MiB	ron.traineddata
rus	Russian	139.0 B	rus.cube.fold
rus	Russian	278.0 B	rus.cube.lm
rus	Russian	891.4 KiB	rus.cube.nn
rus	Russian	317.0 B	rus.cube.params
rus	Russian	14.5 MiB	rus.cube.size
rus	Russian	6.7 MiB	rus.cube.word-freq
rus	Russian	15.4 MiB	rus.traineddata
san	Sanskrit	21.7 MiB	san.traineddata
sin	Sinhala; Sinhalese	6.5 MiB	sin.traineddata
slk	Slovak	8.7 MiB	slk.traineddata
slk_frak	Slovak – Fraktur	825.4 KiB	slk_frak.traineddata
slv	Slovenian	6.5 MiB	slv.traineddata
spa	Spanish; Castilian	128.9 KiB	spa.cube.bigrams
spa	Spanish; Castilian	76.0 B	spa.cube.fold
spa	Spanish; Castilian	248.0 B	spa.cube.lm
spa	Spanish; Castilian	887.5 KiB	spa.cube.nn
spa	Spanish; Castilian	243.0 B	spa.cube.params
spa	Spanish; Castilian	18.1 MiB	spa.cube.size
spa	Spanish; Castilian	3.1 MiB	spa.cube.word-freq
spa	Spanish; Castilian	15.2 MiB	spa.traineddata
spa_old	Spanish; Castilian – Old	16.0 MiB	spa_old.traineddata
sqi	Albanian	6.3 MiB	sqi.traineddata
srp	Serbian	4.4 MiB	srp.traineddata
srp_latn	Serbian – Latin	5.8 MiB	srp_latn.traineddata
swa	Swahili	3.7 MiB	swa.traineddata
swe	Swedish	9.0 MiB	swe.traineddata
syr	Syriac	2.6 MiB	syr.traineddata
tam	Tamil	4.9 MiB	tam.traineddata
tel	Telugu	37.5 MiB	tel.traineddata
tgk	Tajik	1.1 MiB	tgk.traineddata
tgl	Tagalog	3.9 MiB	tgl.traineddata
tha	Thai	12.9 MiB	tha.traineddata
tir	Tigrinya	1.7 MiB	tir.traineddata
tur	Turkish	13.4 MiB	tur.traineddata
uig	Uighur; Uyghur	1.9 MiB	uig.traineddata
ukr	Ukrainian	7.7 MiB	ukr.traineddata
urd	Urdu	4.6 MiB	urd.traineddata
uzb	Uzbek	4.1 MiB	uzb.traineddata
uzb_cyrl	Uzbek – Cyrilic	3.2 MiB	uzb_cyrl.traineddata
vie	Vietnamese	5.8 MiB	vie.traineddata
yid	Yiddish	4.0 MiB	yid.traineddata

© projekt sk-spell