qurator-spk/eynollah
1
0
Fork 0
mirror of https://github.com/qurator-spk/eynollah.git synced 2025-10-26 23:34:13 +01:00
Code Issues Projects Releases Wiki Activity

Commit graph

  • eaff9e3537 Merge branch 'main' into remove-imutils-1 cneud 2025-03-26 20:16:46 +01:00
  • 7df0427b04 In the context of OCR, if Page-XML files already contain text, the new predicted text will replace the existing text. vahidrezanezhad 2025-03-26 18:42:06 +01:00
  • 370d44a66b Slope deskew in the light version is set to zero because when the slope_deskew value exceeds the slope_threshold, the reading order becomes incorrect. This issue needs to be addressed. Additionally, the textlines order within text region in the light version was reversed, and this has been corrected. vahidrezanezhad 2025-03-26 10:45:34 +01:00
  • 4179a2ea3a integrate binarization in ocrd-tool.json cneud 2025-03-26 01:27:02 +01:00
  • 284c2aed7f unresolved references cneud 2025-03-26 01:06:19 +01:00
  • 881f6f37c1 rename ocrd-sbb-binarize to ocrd-eynollah-binarize cneud 2025-03-26 00:50:39 +01:00
  • 3f1e140da1 pep8: class names should use CapWords cneud 2025-03-26 00:32:10 +01:00
  • 62e8d78e73 few minor fixes cneud 2025-03-26 00:27:06 +01:00
  • b4d168cae3 pep8-e265: comment should start with # cneud 2025-03-25 23:54:31 +01:00
  • 87ae6d11a9 pep8: whitespaces around operators cneud 2025-03-25 23:01:12 +01:00
  • fa7bb63481 pep8-e302: two blank lines between functions cneud 2025-03-25 22:36:22 +01:00
  • caf0fbe90f pep8-e302: two blank lines between functions cneud 2025-03-25 22:32:59 +01:00
  • 0e8ea64ba4 don't use equality operators to compare with None cneud 2025-03-25 21:49:35 +01:00
  • 2722a9a464 simplify chained comparisons cneud 2025-03-25 21:44:52 +01:00
  • b9030f5203 replace multi-step list initialization with list literal cneud 2025-03-25 21:42:49 +01:00
  • 5326104d26 remove unnecessary backslash cneud 2025-03-25 21:40:49 +01:00
  • badf80039f remove redundant parentheses cneud 2025-03-25 21:39:00 +01:00
  • 005b6988f4
         Merge pull request #140 from qurator-spk/machine_based_reading_order_integration Clemens Neudecker 2025-03-25 11:00:44 +01:00
  • d3a4c06e7f This commit enables the export of cropped text line images along with their corresponding texts from a Page-XML file. These exported text line images and texts can be utilized for training a text line-based OCR model. vahidrezanezhad 2025-03-20 18:21:44 +01:00
  • 363c343b37 visualising reaidng order- Overlaying on image is provided vahidrezanezhad 2025-03-17 20:09:48 +01:00
  • c8b8529951 For the CNN-RNN OCR model, long text lines are split into two segments vahidrezanezhad 2025-03-17 19:50:58 +01:00
  • 90a1b186f7 this enables to visualize reading order of textregions provided in page-xml files vahidrezanezhad 2025-03-14 17:20:33 +01:00
  • aa72ca3006 Resolved an issue in the OCR-D framework where dir_out received a None value vahidrezanezhad 2025-03-13 15:02:38 +01:00
  • a4f1f35125 Resolving test failure vahidrezanezhad 2025-03-07 13:19:56 +01:00
  • 54040c1db4 Merge remote-tracking branch 'bertsky/machine_based_reading_order_integration_fixes' into machine_based_reading_order_integration kba 2025-03-06 15:48:52 +01:00
  • 0b2c1b9275 remove imutils dependency cneud 2025-03-03 22:21:57 +01:00
  • 687aba1fa2
         replace usages of imutils with opencv equivalents Clemens Neudecker 2025-03-03 22:10:40 +01:00
  • 7110bd971f resolved an error for light version in the case that slope_deskew is smaller than slope_threshold vahidrezanezhad 2025-02-27 19:11:15 +01:00
  • 25116a2c79 resolved 2 errors vahidrezanezhad 2025-02-19 00:35:48 +01:00
  • 869110f185 merge main kba 2025-01-20 14:45:27 +01:00
  • 33fda2f8be changing cnn ocr model name vahidrezanezhad 2024-12-26 22:45:40 +01:00
  • 335aa273a1 simplify, wrap extremely long lines Robert Sachunsky 2024-12-23 03:13:21 +00:00
  • cfc65128b1 reduce redundancy/indentation Robert Sachunsky 2024-12-22 14:56:32 +00:00
  • 01376af905 do_order_of_regions_with_model: simplify Robert Sachunsky 2024-12-22 13:10:05 +00:00
  • 92bfac4b41 Provide OCR as an option to process a directory of XML files, incorporating layout and text line coordinates. vahidrezanezhad 2024-12-20 15:47:21 +01:00
  • fbeef79d50 adding scatter_nd inference vahidrezanezhad 2024-12-16 01:11:54 +01:00
  • 0ae28f7d3e switch from stdlib to loky.ProcessPoolExecutor, ensure shutdown Robert Sachunsky 2024-12-14 12:15:56 +00:00
  • f93c6c288d function of patch-wise inference with scatter_nd is added vahidrezanezhad 2024-12-14 02:50:17 +01:00
  • 0e8c561618 debugging issues vahidrezanezhad 2024-12-14 00:24:29 +01:00
  • e9c0d716f6 CI: install optional dependencies, too Robert Sachunsky 2024-12-11 23:48:56 +00:00
  • dcaf796283 change polarity of orientation angle (PAGE schema required cw=positive) Robert Sachunsky 2024-12-11 23:07:56 +00:00
  • b4b0890294 add option to overwrite output xml, but skip by default if file exists Robert Sachunsky 2024-12-11 18:45:18 +00:00
  • b9ca7a6191 log num_cols-dependent resizing Robert Sachunsky 2024-12-11 18:44:54 +00:00
  • 9270ea4550 annotate region angles in PAGE Robert Sachunsky 2024-12-11 18:37:20 +00:00
  • 3b70b11ea6 avoid deskewing patches if binary-empty Robert Sachunsky 2024-12-11 18:36:20 +00:00
  • 7e9ee90e6e switch from (ad-hoc) mp.Pool to (attribute) concurrent.futures.ProcessPoolExecutor Robert Sachunsky 2024-12-11 12:18:29 +00:00
  • 68456ea002 do_work_of_slopes_new*, do_back_rotation_and_get_cnt_back, do_work_of_contours_in_image: use mp.Pool, simplify Robert Sachunsky 2024-12-11 11:30:38 +00:00
  • 25e967397d exit early if no text regions found (to avoid segfault) Robert Sachunsky 2024-12-11 11:24:56 +00:00
  • 21efea8711 no del on function argument Robert Sachunsky 2024-12-11 18:36:57 +00:00
  • 5e0c1da711 simplify Robert Sachunsky 2024-12-11 00:18:58 +00:00
  • 54cb15056b do_image_rotation / return_deskew_slop: avoid code duplication, simplify via mp.Pool Robert Sachunsky 2024-12-09 16:37:34 +00:00
  • 6fe02df973 do_image_rotation: fix f93fa12 (do return results) Robert Sachunsky 2024-12-09 16:35:31 +00:00
  • d68017037c do_prediction: trigger GC to avoid CUDA OOM Robert Sachunsky 2024-12-09 11:27:11 +00:00
  • ad748d0039 do_prediction: avoid code duplication Robert Sachunsky 2024-12-09 10:55:41 +00:00
  • c3163caefd avoid indentation Robert Sachunsky 2024-12-05 14:28:17 +00:00
  • 055463d23a avoid indentation Robert Sachunsky 2024-12-05 09:43:30 +00:00
  • aaea2ef463 simplify Robert Sachunsky 2024-12-05 09:40:02 +00:00
  • 3d88b207fc run: log instead of print Robert Sachunsky 2024-12-05 09:39:55 +00:00
  • a520bd1f77 wrap extremely long lines Robert Sachunsky 2024-12-04 22:49:34 +00:00
  • cd4e426977 avoid indentation (skip_layout_and_reading_order) Robert Sachunsky 2024-12-04 22:11:34 +00:00
  • 5b82320707 avoid indentation Robert Sachunsky 2024-12-04 22:09:32 +00:00
  • 9f12fa241d log-level: only set 'eynollah' logger level Robert Sachunsky 2024-12-04 22:09:15 +00:00
  • 14beb46224 simplify loading models w/o dir_in mode Robert Sachunsky 2024-12-04 21:07:26 +00:00
  • 329fac23f6 do not reload enhancement model in dir_in mode, simplify Robert Sachunsky 2024-12-04 18:29:49 +00:00
  • 3b9a29bc5c simplify dir_in conditionals Robert Sachunsky 2024-12-04 18:19:54 +00:00
  • 7ae64f3717 RO model: do not reload when in dir_in mode Robert Sachunsky 2024-12-04 16:18:35 +00:00
  • f765e2603b move Torch to optional dependencies (to avoid clash with TF over CuDNN) Robert Sachunsky 2024-12-04 15:57:13 +00:00
  • 871d7bfc5a fixed: machine based reading order cause tuple index out of range error if number of textregion is one. vahidrezanezhad 2024-12-04 16:41:00 +01:00
  • 6aad006f4c filter textregions without textline vahidrezanezhad 2024-12-02 12:43:57 +01:00
  • 1083d1c7fb gha: try to free disk space kba 2024-11-25 19:32:42 +01:00
  • 8014a9e416
         Update Makefile vahidrezanezhad 2024-11-22 19:47:06 +01:00
  • 3000255a24
         Update Makefile vahidrezanezhad 2024-11-22 12:40:21 +01:00
  • 1746920275
         Update Makefile vahidrezanezhad 2024-11-21 12:08:29 +01:00
  • b622494f34 new table detection model is integrated vahidrezanezhad 2024-11-21 02:16:22 +01:00
  • e9b860b275 artificial_class_label for table region vahidrezanezhad 2024-11-18 16:34:53 +01:00
  • d9f79c3404 fixing IndexError by reading order detection vahidrezanezhad 2024-11-18 10:15:19 +01:00
  • 5fa8ca46a4 updating requirements vahidrezanezhad 2024-11-14 17:35:00 +01:00
  • ce5b611296 tests are passed - new models by the way should be uploaded vahidrezanezhad 2024-11-14 17:18:07 +01:00
  • 238ea3bd8e update resizing in inference vahidrezanezhad 2024-11-14 16:26:19 +01:00
  • f43c49c508 textlines of drop capitals are connected to corresponding textline if possible otherwise they are inserted in corresponding textregion vahidrezanezhad 2024-11-13 11:53:56 +01:00
  • 22b0b07a73 drop capital and marginals extraction is updated vahidrezanezhad 2024-11-11 19:01:40 +01:00
  • 1ae77e61c8
         Update requirements.txt Clemens Neudecker 2024-11-11 14:11:36 +01:00
  • 8409de0e58 sbb_binarization is integrated into eynollah works in framework of ocrd - sbb_binarization in ocrd works for individual images by the way as standalone flowing from directory can be used now. For eynollah in ocrd framework I have added -light version as default parameter. vahidrezanezhad 2024-11-10 19:34:43 +01:00
  • 0914b5ff8a resolve merge conflict of main branch with machine based reading order branch vahidrezanezhad 2024-11-06 00:34:00 +01:00
  • 6aee70d0cd Resolve merge conflict of main and machine based reading order branch vahidrezanezhad 2024-11-06 00:10:25 +01:00
  • bceeeb56c1
         Merge pull request #138 from qurator-spk/extracting_images_only vahidrezanezhad 2024-11-05 22:10:51 +01:00
  • f7e5fb917f resolving merge conflict of machine based reading order and extracting only images branches vahidrezanezhad 2024-11-05 22:09:39 +01:00
  • 751b0102f7 updating early layout inference for light version vahidrezanezhad 2024-11-05 19:50:18 +01:00
  • e796a99c5c updating inference for early layout in the case of documents with number of columns bigger than 2 vahidrezanezhad 2024-10-30 15:02:50 +01:00
  • 438df52287 updating vahidrezanezhad 2024-10-30 00:52:09 +01:00
  • 7b4d14b19f addinh shifting augmentation vahidrezanezhad 2024-10-29 17:06:22 +01:00
  • 90ee2d61dc textline segmentation is masked with drop capitals vahidrezanezhad 2024-10-28 20:56:06 +01:00
  • 5037e9896d Merge branch 'machine_based_reading_order_integration' of https://github.com/qurator-spk/eynollah into machine_based_reading_order_integration vahidrezanezhad 2024-10-25 19:47:20 +02:00
  • 82281bd6cf fixing a bug occuring with reading order + Slro option with no patch textline model and thresholding artificial class vahidrezanezhad 2024-10-25 19:42:48 +02:00
  • fd14e656aa early_erosion is added vahidrezanezhad 2024-10-25 14:01:39 +02:00
  • 328d33e3dc Temporary commit – textline prediction without patches vahidrezanezhad 2024-10-23 16:55:41 +02:00
  • 70772d4104 binarization as a standalone command vahidrezanezhad 2024-10-21 23:46:38 +02:00
  • 451188c3b9
         Changed deprecated lr to learning_rate and model.fit_generator to model.fit johnlockejrr 2024-10-19 13:25:50 -07:00
  • f09eed1197 Changed deprecated lr to learning_rate and model.fit_generator to model.fit johnlockejrr 2024-10-19 13:25:50 -07:00
  • df4a47ae6f
         Update inference.py to check if save_layout was passed as argument otherwise can give an cv2 error johnlockejrr 2024-10-19 13:21:29 -07:00