1774076f4a
updating light version. Remove textlines or textregion contours inside a bigger one
3 months ago
ad32316217
updating light version
3 months ago
133091137d
dilation of textregions and marginals are accomplished
3 months ago
95effe54a0
updating textregions dilation
3 months ago
b33739adee
parametriyation in the case of textline contours dilation is accomplished
3 months ago
6626dc6866
updating textline dilation parameters
3 months ago
62f8ae4860
updating dilation of textlines and text regions
3 months ago
7f08458436
dilation of text regions without opencv
3 months ago
5d680136a4
updating light version
3 months ago
b9e8959c4a
update of light versions
3 months ago
2d18739d9b
postprocessing of textline contour dilation + skip layout and reading order passed as an argument
3 months ago
5a07cd9cfa
the most effective version of contours dilation without opencv and all at once
3 months ago
a1f1f98de3
updating scaling contours
3 months ago
21380fc870
scaling contours without dilation
3 months ago
1b18ae874b
passing number of columns as an argument
3 months ago
2c93904985
avoiding double binarization
3 months ago
f0b49073b7
adding option for textline detection in printspace
4 months ago
c3a4a1bba7
resolving issue #110 in a better way
4 months ago
0f87974b0c
writing drop capitals in xml output + and may resolve issue #110
4 months ago
93005959e5
inference batch size debugged
4 months ago
7ae6a8776f
ignoring dpi check by light version
4 months ago
04e79002b3
making light version faster for 1 and 2 columns images
4 months ago
c10a525675
inference with batch size bigger than 1
4 months ago
4f8210de71
update Makefile model location
4 months ago
6f4205ba49
update pyproject.toml
4 months ago
74eac4dacc
dtype = object in the case of length 1 arise error
4 months ago
4c50479cb8
pyproject.toml may work for ocrd
4 months ago
53fd5fb2a5
resolving #106 for pyproject.toml test
4 months ago
e976778796
testing pyproject.toml
4 months ago
00bf2b64d0
1&2 column images only printspace
5 months ago
be144db9f8
updating 1&2 columns images + full layout
5 months ago
a62ae370c3
new full layout model and early layout for 1&2 column images are integrated - light version
5 months ago
5144668834
ocr engine first integration
5 months ago
eac18c553d
machine based reading order as an argument
1 year ago
941d87328a
machine based reading order & works for not full layout case
1 year ago
59c0d90e5a
machine based reading order inference & optimized algorithm
1 year ago
49c93149a4
machine based reading order inference with a variable batch size
1 year ago
5fdc6d4fa4
integration of machine based reading order detection
1 year ago
fc9e9cc29f
Merge pull request #117 from qurator-spk/tf-2.12-or-greater
...
Update tensorflow
1 year ago
4254ce3bdb
Update README.md
1 year ago
56934c876a
remove duplicate test for Python 3.8
1 year ago
6c65fc4dfe
Update config.yml
1 year ago
9d3a1a5b76
Update test-eynollah.yml
1 year ago
03bfd7a390
Update requirements.txt
...
Update to `tensorflow>=2.12` (drops Python 3.7 support)
* fix #114
* fix #115
Tested by @vahidrezanezhad @cneud
1 year ago
7a70b20c77
apply missed commit #a56988a back
1 year ago
0ea90b7509
Merge pull request #116 from qurator-spk/fix-typos
...
fix some typos
1 year ago
830b0fc8fa
Merge pull request #113 from qurator-spk/tf_<2.12.0
...
introduce (temporary) upper bounds for tensorflow and numpy version
1 year ago
d3b06baa84
fix some typos
1 year ago
3e62657570
added the TF upper bound in the README
1 year ago
b58a327c5d
cap numpy to <1.24.0
...
OK so now numpy is the culprit (shipped unbound via ocrd) which had several deprecations expire with release of v1.24.0 that require changes to our codebase, e.g.
* The deprecation for the aliases np.object, np.bool, np.float, np.complex, np.str, and np.int is expired
* Ragged array creation will now always raise a ValueError unless dtype=object is passed.
See also here: https://numpy.org/devdocs/release/1.24.0-notes.html#expired-deprecations
1 year ago
vahidrezanezhad
cneud