38bf0d8740
solving issue by loading model by directory as input
2 years ago
583cdcee2c
new (hybrid cnn+transformer) textline model which can accelerate to extract contour textlines faster
2 years ago
8d5079c909
issue #77 is resolved on main branch
2 years ago
402c5339ac
issue #77 is resolved
2 years ago
01bfc3914d
extracting page as an option
3 years ago
ae7c424889
Update eynollah.py
3 years ago
cd9920eea7
extracting page
3 years ago
735abc43f3
option to ignore page extraction
3 years ago
934bbd5892
cleanup
3 years ago
ecf117ca95
adapt to tf1.compat session mode in tf2
3 years ago
568391ec4a
require model command line option ( fix #59 ) ( #73 )
3 years ago
3bbbeecfec
all options are enabled for light version
3 years ago
Gerber, Mike
f27ac155ae
🧹 Downgrade "Patch size" log message to debug
...
Fixes gh-55.
3 years ago
adf10942fa
issue #55 resolved
3 years ago
2eacb9a8ec
renaming the models
3 years ago
c606391c31
flow from directory
3 years ago
cf5ef8f5ae
light version as option
3 years ago
b8a532180a
light version integration
3 years ago
10f1acef29
Merge pull request #65 from mikegerber/fix/enhanced-message
...
Fix/enhanced message
3 years ago
c30d4d5c30
Merge pull request #64 from mikegerber/feat/better-time-msgs
...
💄 Improve timing messages (Fixes #62 )
3 years ago
Gerber, Mike
11d9b00510
🧹 Don't produce spurious TextEquiv elements.
...
eynollah produces spurious - and empy - pcGts TextEquiv elements. This
is a. unnecessary, b. wrong and c. produces a lot of warning messages
in subsequent OCR processing steps because the OCR processor warns
about already existing text.
Fix this by not generating any TextEquiv elements.
Fixes gh-37.
3 years ago
Gerber, Mike
1fe8f92afc
🐛 Clarify message if an image was enhanced
3 years ago
Gerber, Mike
7ccd7663e1
💄 Improve more timing messages
3 years ago
Gerber, Mike
cdea0acffe
💄 Improve timing messages ( Fixes #62 )
3 years ago
f0ac0bb090
📦 v0.0.11
3 years ago
d75803b11d
ocrd-tool: "models" parameter is a directory
3 years ago
e769f625fe
📦 v0.0.10
3 years ago
09d85bee87
Merge remote-tracking branch 'vahidrezanezhad/main' into main
3 years ago
169b50aaaf
fixed: empty page error due None table contours
3 years ago
0e63ebcbe5
📦 v0.0.9
3 years ago
4223fed628
Merge remote-tracking branch 'vahidrezanezhad/main' into main
3 years ago
e7868b9851
📦 v0.0.8
3 years ago
5124a60527
set pcGtsId before adding file to mets
3 years ago
0859d22f4c
modifications
3 years ago
14c588e162
resolving an issue
3 years ago
254abf4d3d
more modifications for tables
4 years ago
b3b49272a5
README is updated
4 years ago
c67e155431
table detection completed, enhanced images can be now written to output
4 years ago
a5c940705a
tables are integrated
4 years ago
80b17af40c
#47 fixed
4 years ago
d784202ae1
📦 v0.0.7
4 years ago
6b810eb682
Merge remote-tracking branch 'vahidrezanezhad/main' into main
4 years ago
4560738427
#45 fixed
4 years ago
efc146feb8
📦 v0.0.6
4 years ago
becb0c1329
trivial
4 years ago
059905c9e4
#43 empty textlines caused by newer python-opencv, is resolved
4 years ago
d1330ffb80
#43 resolved
4 years ago
80795c9e6c
📦 v0.0.5
4 years ago
45939abdff
OCR-D CLI: remove allow_enhancement parameter
...
It does not toggle enhancement (eynollah does that internally anyway)
but setting it to true will base the coordinate calculations on that
enhanced (different-sized) image instead of the original. That is never
sensible in the OCR-D context.
4 years ago
5d2fe79822
📦 v0.0.4
4 years ago
43c9302390
fixed #40 and separators are also written in xml
4 years ago
fce7cdfd8b
📦 v0.0.3
4 years ago
aa2e91641a
Merge branch 'main' of https://github.com/qurator-spk/eynollah into main
4 years ago
799a7c7632
fixed #38
4 years ago
26283c6a3b
📦 v0.0.2
4 years ago
c4b2c71e68
resolving issue https://github.com/qurator-spk/eynollah/issues/38
4 years ago
7cbecadccc
adding the binarization model and option to binarize input document for the cases like dark, stronly bright and other ones
4 years ago
44dad6a072
strong erosion, more modification
4 years ago
176c7531ab
Update eynollah.py
4 years ago
c051e22432
fixing again the error raised because of erosion
4 years ago
d5be8aece3
Merge pull request #33 from qurator-spk/ocrd-cli
...
Ocrd cli
4 years ago
6c8852eb04
check_dpi: catch Pillow choking on faulty img, return 230
4 years ago
ff265eee5c
cv2pil: do COLOR_BGR2RGB conversion
4 years ago
c7f304dcb6
ocrd processor: pass local filename as image_filename, ht @bertsky
4 years ago
d0b0e23ac6
do DPI calculation as part of caching images
4 years ago
ae0b4a825a
ocrd cli: catch dpi == 1, return 230
4 years ago
2e8a3e3bee
use Page.imageFilename directly for accurate DPI estimate
4 years ago
42ccb4711d
Update qurator/eynollah/ocrd-tool.json
...
Co-authored-by: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
4 years ago
1184d3d2fc
issue raised by Clemens, strong erosion causing
4 years ago
4897cefdb7
allow passing PIL image to Eynollah w/o disk I/O
4 years ago
d40c453dad
check_dpi: raise exception if resolution == 1 to trigger except clause
...
Co-authored-by: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
4 years ago
1367f82605
improve ocrd-tool descriptions
...
Co-authored-by: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
4 years ago
037210b292
update writer.py
4 years ago
b8d818ede1
writer: don't create empty PcGts at init
4 years ago
8c4e9b6068
allow passing pcgts to eynollah and writer
4 years ago
2bc34891a5
fix CLI call
4 years ago
98f9272c4b
a trivial issue is resolved
4 years ago
9db6edf51e
OCR-D CLI
4 years ago
1715f0d8b3
allow overriding DPI
4 years ago
c95529725a
🐛 typo type{,_}
4 years ago
93f93444ee
🐛 typo {c,C}oords
4 years ago
416a84e542
replace lxml with OCR-D/core PAGE API
4 years ago
7a859ffae4
Merge branch 'main' into xml-rfct
4 years ago
d5a9817390
back on track- freezing problem , memory error and issues with reading order by drop capitals and marginals are resolved
4 years ago
43b8759acf
back on track- freezing problem , memory error and issues with reading order by drop capitals and marginals are resolved
4 years ago
b473c85a59
OOM error happend with tensorflow-gpu=1.15.5 is resolved
4 years ago
3d9da4feaa
writer: use a single counter for all regions/lines
4 years ago
a678bbf966
counter: add reset();
4 years ago
a3465ca1a0
eliminate id_of_texts from xml_reading_order, fix plus one error
4 years ago
6c60d9e90a
reading order: fix @index
4 years ago
02aa31cc66
Merge remote-tracking branch 'origin/main' into xml-rfct
4 years ago
c5736e9b74
fix region counting
4 years ago
67a9fc8820
..
4 years ago
4b3c8a6707
bug in reading order is fixed
4 years ago
73b7c780ab
Update eynollah.py
...
reading order bug for documents with text regions less than 5: fixed
4 years ago
03d75f5788
simplify serialize_lines_in_region
4 years ago
d95fcf14c0
id_of_marginalia still necessary
4 years ago
56b688befe
counter: allow arbitrary line/region id
4 years ago
7eb973b3aa
xml_reading_order takes id_of_marginals directly
4 years ago
98568402c7
counter: init-overrideable
4 years ago
vahid
vahidrezanezhad
cneud
Clemens Neudecker
Gerber, Mike
Konstantin Baierer
vahid
Konstantin Baierer
vahid
vahid