Commit Graph

3979 Commits

Author SHA1 Message Date
Armyke
25fa392887 Added an additional optional --tmp_dir parameter to specify the temporary directory in which tesstrain.py creates the training temporary files. The main reason is due to the slow R/W on HDD, if anyone wants to speed up this process can use as tmp_dir a directory on an SSDrive 2019-03-04 13:26:53 +00:00
Stefan Weil
7fbde96a04 Format new code with clang-format
Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-03-02 20:26:07 +01:00
Stefan Weil
38fac625cd Format new code with clang-format
Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-03-02 20:01:48 +01:00
Shree
a0202bac70 Rename function to TessBaseAPIGetTsvText to be consistent to the Create method 2019-03-02 16:29:53 +00:00
zdenop
5de2a21b3f
Merge pull request #2283 from Shreeshrii/lstmbox
Add missing renderers to C-API
2019-03-02 15:15:34 +01:00
zdenop
198c90b124
Merge pull request #2285 from stweil/opt
PAGE_RES_IT: Optimize compare operators by using inline code
2019-03-02 15:13:14 +01:00
Stefan Weil
9c90894ff0 PAGE_RES_IT: Optimize compare operators by using inline code
Avoiding a function call will make both == and != operator faster.

Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-03-02 14:57:16 +01:00
Egor Pugin
7cc97c25ca
Merge pull request #2284 from stweil/fix
Fix some compiler warnings
2019-03-02 16:35:55 +03:00
Stefan Weil
295996ed05 commandlineflags: Fix compiler warnings (signed/unsigned)
Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-03-02 14:21:04 +01:00
Stefan Weil
eb14726aac ICOORD: Fix old type casts
This fixes compiler warnings and avoids unnecessary conversions
between float and double.

Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-03-02 14:04:54 +01:00
Stefan Weil
fb0f1bcf66 BoxChar: Fix compiler warnings (signed/unsigned)
Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-03-02 14:04:54 +01:00
Stefan Weil
0e1a1fc3cf Validator: Fix compiler warnings (signed/unsigned)
This also fixes a regression in validate_grapheme_test introduced
by commit 32e9d7c8f5.

Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-03-02 13:05:03 +01:00
Shree
c7e8131efc Add TSV option to C-API 2019-03-02 09:50:54 +00:00
Shree
22c099348b rename LSTMBOX to LSTMBox 2019-03-02 09:11:47 +00:00
zdenop
f5a7ca26e7
Merge pull request #2244 from Shreeshrii/mya
Fix Myanmar validation rules as per Unicode charts
2019-03-01 18:37:36 +01:00
zdenop
2ba8e0061a
Merge branch 'master' into mya 2019-03-01 18:37:24 +01:00
zdenop
0b354f2b84
Merge pull request #2282 from Shreeshrii/configs
Add lstmbox and wordstrbox to C-API
2019-03-01 18:33:29 +01:00
Shree
c33f03e33e Add lstmboxand wordstrbox to capi.h 2019-03-01 17:16:59 +00:00
Shree
76ec21df3d Add lstmbox and wordstrbox to C-API 2019-03-01 16:40:41 +00:00
zdenop
c4b5178296
Merge pull request #2280 from Shreeshrii/configs
install lstmbox and wordstrbox config files
2019-03-01 17:20:55 +01:00
Shree
08e96516c6 install lstmbox and wordstrbox config files 2019-03-01 15:26:59 +00:00
zdenop
a783009189
Merge pull request #2279 from Shreeshrii/tamil_numbers
correct handling of 0BF0-0BFA Tamil numbers and symbols
2019-03-01 14:36:20 +01:00
zdenop
646b043d2c
use space instead of tab 2019-03-01 14:36:09 +01:00
Shree
5ee1deaea2 correct handling of 0BF0-0BFA Tamil numbers and symbols 2019-03-01 13:21:49 +00:00
zdenop
d7ddc4c5b7
Merge pull request #2270 from Shreeshrii/U_ARABIC_NUMBER
Treat U_ARABIC_NUMBER as LTR
2019-02-28 09:27:54 +01:00
zdenop
2a69a4b4e1
Merge pull request #2275 from russiaayya/patch-1
Change option -l to --lang
2019-02-27 20:22:32 +01:00
russiaayya
c6cc54aa76
Change option -l to --lang 2019-02-27 12:55:34 -05:00
zdenop
12c1225a5f
Merge pull request #2271 from stweil/refactor
Refactor class Network
2019-02-27 07:43:13 +01:00
Stefan Weil
13bd96fd37
Merge pull request #2272 from nijel/patch-1
Allow UTF-8 variant of C locale
2019-02-26 22:25:48 +01:00
Michal Čihař
14c4494f42 Allow UTF-8 variant of C locale
It behaves same in scanf, but it allows proper handling of unicode
chars.
2019-02-26 21:37:33 +01:00
Stefan Weil
98dd3b6351 Refactor class Network
That class is an abstract class with several pure virtual functions.

Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-02-26 16:55:31 +01:00
zdenop
cf85054453
Merge pull request #2269 from Shreeshrii/grapheme
Use alternative way to comment a block of code
2019-02-26 16:51:34 +01:00
Shree
25b02bf1f2 Treat U_ARABIC_NUMBER as LTR 2019-02-26 09:51:21 +00:00
Shreeshrii
2f71fe280c
Use alternative way to comment a block of code (using the c preprocessor).
https://github.com/tesseract-ocr/tesseract/pull/2268#pullrequestreview-207605382
Thanks @amitdo
2019-02-26 15:05:51 +05:30
zdenop
9ddf267907
Merge pull request #2268 from Shreeshrii/grapheme
Remove test for Word started with a combiner
2019-02-25 20:59:40 +01:00
Shree
449f1cd4ba Remove test for Word started with a combiner 2019-02-25 18:47:42 +00:00
zdenop
8e83b209c6
Merge pull request #2262 from Shreeshrii/distort
Fix PrepareDistortedPix based on feedback
2019-02-23 18:23:24 +01:00
zdenop
25c43b1e7c
Merge branch 'master' into distort 2019-02-23 18:23:14 +01:00
Stefan Weil
b3e355a682 Remove whitespace at line endings
Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-02-23 17:49:56 +01:00
Shreeshrii
34e4d6b1d7
Revert to 0 (50% percents of images inverted). 2019-02-23 17:59:00 +05:30
Shreeshrii
287d5341bf
TODO 2019-02-23 17:56:02 +05:30
Shreeshrii
3e3e1ed55d
Remove commented Code 2019-02-23 17:54:00 +05:30
zdenop
c02f5e99fc
Merge pull request #2259 from Shreeshrii/distort
implement PrepareDistortedPix as part of DegradeImage
2019-02-22 21:06:29 +01:00
Shree
2aded47a3c Implement distort_image in text2image - default false 2019-02-22 12:27:27 +00:00
Shree
49ed3a72d4 implement PrepareDistortedPix as part of DegradeImage 2019-02-21 14:48:29 +00:00
zdenop
e250f3422d
Merge pull request #2258 from stweil/doc
Fix doxygen comments
2019-02-21 07:41:22 +01:00
zdenop
b6868599dc
Merge pull request #2257 from stweil/dp
Optimize calculation of dot product for double vectors with AVX
2019-02-21 07:40:22 +01:00
Egor Pugin
9644b582fe
Update appveyor.yml 2019-02-21 01:23:50 +03:00
Stefan Weil
2cbe723d03 Fix doxygen comments
Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-02-20 21:11:38 +01:00
Stefan Weil
ef4d5b2e69 Optimize calculation of dot product for double vectors with AVX
This improves the performance with best models and should also
make training faster.

Signed-off-by: Stefan Weil <sw@weilnetz.de>
2019-02-20 17:45:38 +01:00