Releases: OCR-D/ocrd_all
Releases · OCR-D/ocrd_all
v2020-12-28
ocrd_cis c4d0fae..925d744
Release: v0.1.5
- traverse regions in reading-order (so derived images are, too)
- resegment: more robust concave hull (really)
- resegment: more robust concave hull
- segment: skip segments with zero height or width
- deskew/binarize/denoise: skip segments with zero height or width
- segment: don't hmerge across separators
- fix logging typo in 75f185bb1
- segment: improve over 62a96f91e by re-ordering, not only grouping ignored and new regions
ocrd_fileformat 04e4663..ae52a2e
Release: v0.1.2
- 📦 v0.1.2
- Merge pull request #24 from OCR-D/exit-gracefully
ocrd_tesserocr 6ab6167..89b73ff
Release: v0.10.1
ocrd_typegroups_classifier dc2f685..54c8a78
Release: v0.2.0
opencv-python f33e65d..bbaa777
Release: 46
- Merge pull request #428 from ApproximateIdentity/debug-build-documentation
sbb_textline_detector 17069f7..4c498fc
- resolving issue qurator-spk/sbb_textline_detection#53
tesseract 66175f97..190be30a
Release: 5.0.0-alpha-20201224
- Create new pre-release 5.0.0-alpha-20201224
- Fix dependency on tmp directory for unittest programs
- Fix tar command for variants which require -j or -z
- Simplify code for equationdetect_test
- Simplify code for cleanapi_test
- Remove unwanted # at EOL
- Merge pull request #3191 from stweil/unittest
- Merge pull request #3190 from stweil/pango_font_info_test
- Merge pull request #3189 from stweil/pango_font_info_test
- Fix configure for flat training build without asciidoc
- Fix installation of training tools for flat training build
- Add XML_CATALOG_FILES for MacOS with Homebrew (#3188)
- Fix unittest for flag training build
- Merge branch 'master' of github.com-egorpugin:tesseract-ocr/tesseract
- Add include headers to cmake target.
- Fix git submodule 'test'
- Merge pull request #3179 from Shatur95/fix-cmake-issues
- Merge pull request #3181 from stweil/m1
v2020-12-10
core 2298662..135acb6
Release: v2.21.0
- 📝 changelog
- Merge pull request #651 from OCR-D/bertsky-fix-task-sequence
ocrd_cis 36f4a95..c4d0fae
Release: v0.1.5
- resegment: improve over new polygonal/global algorithm…
- resegment: complete rewrite (polygonal and global)…
- common: in hmerge_line_seeds, find and fill shortest gap…
ocrd_segment 907ba7d..a26a3c3
Release: v0.1.6
- repair: also ensure >2 points
ocrd_tesserocr 24b7ced..6ab6167
Release: v0.10.1
ocrd_typegroups_classifier 3d0004f..dc2f685
Release: v0.1.3
- 📦 v0.1.3
- 📝 changelog
- ocrd-tool: add output_file_grp
- Merge branch 'pr/4'
opencv-python a5fc295..f33e65d
Release: 46
- OpenCV 4.5.0
- update OpenCV licensing
tesseract 4c35f51..66175f9
Release: 5.0.0-alpha
- Merge pull request #3172 from MerlijnWajer/master
- Merge pull request #3171 from stweil/lsan
- Fix build with enabled TensorFlow
- Fix regression in Network::Serialize (fix issue #3167)
- Merge pull request #3165 from MerlijnWajer/master
- Merge pull request #3128 from acoder77/patch-1
- Merge pull request #3160 from stweil/string2
- Merge pull request #3159 from stweil/pack
workflow-configuration 2336cbb..3909b7b
- add
make serverbridge forocrd workflow server- make
make showwork independent of workspaces- add show pseudo-target, improve documentation
v2020-11-27
core 079aace..2298662
Release: v2.21.0
- 📦 v2.21.0
- 📝 changelog
- Merge branch 'profile-cpu-time'
- Merge branch 'numpy-any-version'
- 📝 changelog
- Merge remote-tracking branch 'origin/bertsky-patch-task-sequence'
- 📝 changelog
- regenerate PAGE API
- prune_ReadingOrder: also check empty groups
- 📦 v2.20.2
- Merge pull request #645 from OCR-D/fix-644
dinglehopper 389e253..8cd8314
- 🐛 dinglehopper: Bump up ocrd req for zip_input_files
- Merge pull request #49 from kba/zip_input_files
ocrd_calamari 795826f..1c7fcda
Release: v1.0.0
- 📝 README: Link to model download
- 📦 v1.0.0
- 🐛 Pin h5py to < 3 because pip
- Merge branch 'feat/update-calamari1'
ocrd_cis 1c01978..36f4a95
Release: v0.1.5
- segment (page/table level): fix significance criterion for partitions' line labels
- segment (page/table level): continue more gracefully when XY-cut fails
- segment: prevent empty ReadingOrder group
- ensure valid polygons for new coords
- polygon_for_parent: ensure path validity before checking consistency
- segment/common/ocrolib: use better reconstruction…
- segment: fix hmerge transitivity rules
- segment (page level): hmerge line labels (within each region) here, too
- 🔥 morph.label: fix ncomps (+1)
- dewarp: expose smoothness parameter
- deskew: delegate to core for reflection and rotation
- segment: fix pageid
- segment: upgrade segmentation failures from warning to error
- segment (region level): fix and speed up horizontal merging
- segment (page/table level): improve incremental annotation…
- recognize: fix regression in 4f3947d28c
- clip: require independence instead of min_fraction threshold
- clip: avoid suppressing overlapping components from both sides
- segment (region level): annotate clipped region images, too
- segment (region level): do not ignore seps/neighbours after clipping
- resegment: fix against overflow in numpy slices due to rounding errors when cropping
- resegment: use set_points to ensure invalidating existing line images
- ocrolib.spread_labels: fix when no labels exist
- segment: do not suppress neighbours encompassing the full image
ocrd_segment 8b6591e..907ba7d
Release: v0.1.6
- 📦 0.1.6
- replace-original: only ensure coord validity, not consistency
- replace-original: also replace @width/height for clarity
- replace-original: invalidate Border/orientation, add new original as derived image too
- update changelog
- repair: also fix negative coords, also on page level
sbb_binarization 30a3c98..ea69494
Release: v0.0.6
sbb_textline_detector 5d7c090..17069f7
- Merge pull request #51 from qurator-spk/h5py3
tesseract e20ffdd..4c35f51
Release: 5.0.0-alpha
- Merge pull request #3158 from Shreeshrii/master
- Merge pull request #3155 from Shatur95/fix-cmake-targets-path
workflow-configuration a9a797e..2336cbb
- ocrd-import: fix mkdir in workspace dir
- when in a GPU semaphore, don't buffer (log) output
- ocrd-import: ensure OCR-D-IMG exists before convert
- ocrd-import: prefix PAGE-XML file ID by fileGrp in -P mode, too
v2020-11-18
Fixed:
- sub-venv rules: move core dependency to inner make exclusivel, #217
cor-asv-ann 0ae6867..be73c96
Release: v0.1.6
- evaluate: delegate to new zip_input_files in core
core 6ca42f1..079aace
Release: v2.20.1
- 📦 v2.20.1
- 📝 changelog
- ocrd_mets.find: ignore mets:file without mets:FLocat
- workspace.image_from_{page,segment}: warn of duplicate features in AlternativeImages
- processor.base: improve debug log msg
- workspace.image_from_{page,segment}: no Region/@orientation means inherit…
- workspace.image_from_{page,segment}: re-crop after rotation…
- workspace.image_from_{page,segment}: refactor, less redundant warnings…
- 📦 v2.20.0
- 📝 changelog
- Merge pull request #639 from bertsky/fix-page-invalidate-alternativeimages
- 📝 changelog
- Merge pull request #635 from bertsky/processor.zip_input_files
- 📝 changelog
- run_cli: do not capute STDOUT/STDERR
- 📝 changelog
- Merge pull request #637 from bertsky/add-metadata-version
- 📝 changelog
- Merge pull request #636 from b2m/fix-ocrd-utils-on-windows
- 📦 v2.19.0
- 📝 changelog
- Merge pull request #627 from bertsky/fix-pillow-high-depth-images
- 📝 changelog
- Merge pull request #625 from OCR-D/reset-mets-permissions
- Merge pull request #633 from OCR-D/altimage-wo-comment
- 📝 changelog
- Merge pull request #629 from Witiko/master
dinglehopper 6e47acd..389e253
- 🐛 dinglehopper: Fix alto_extract_lines()'s type annotation
- 🐛 dinglehopper: Fix alto_extract()'s type annotation
- ✔️ dinglehopper: Add missing integration test markers
- 📝 dinglehopper: README-DEV: Massage markdown a bit
- Merge pull request #46 from b2m/tool-changes
- 🐛 dinglehopper: Normalize in plain_extract()
- 🎨 dinglehopper: Reformat using black
- 🎨 dinglehopper: s/LOG/log
- Merge pull request #43 from bertsky/patch-1
- Merge pull request #42 from b2m/test-python-cache-for-travis
- Merge pull request #37 from b2m/fix-sort-with-none
- Merge pull request #39 from b2m/test-python-3.9
- Merge pull request #38 from b2m/add-editorconfig
Submodule ocrd_anybaseocr cb82aad...de27915:
ocrd_cis 5ec0e34..1c01978
Release: v0.1.5
- Merge remote-tracking branch 'upstream/master' into fix-resegment
- segment (region level): ignore separators and other existing regions
- resegment: set correct pageId (not pcGtsId or input file ID)
- dewarp: skip zero size (avoid running into numpy conversion problems)
- segment: set pageId for derived images, too
- segment: fix reference before assignment when partitioning
- resegment: skip empty line polygons
ocrd_fileformat ded152d..04e4663
Release: v0.1.1
- 📝 README: fix 2nd workflow-configuration example, fix #22
ocrd_segment aadbc7b..8b6591e
Release: v0.1.5
- 📦 0.1.5
- evaluate: delegate to zip_input_files now in core
- replace_original: delegate to repair.ensure_consistent
- replace_page: delegate to zip_input_files now in core
- replace_page: delegate to repair.ensure_consistent
- replace_page: catch invalid coords and skip page
- replace_page: new OCR-D CLI (inverse of replace_original)
- 📦 0.1.4
- repair: fix coordinate consistency/validity errors if possible
- extract-regions: set pageId for derived images, too
ocrd_wrap 1876b65..9dfbdb3
Release: v0.1.3
- 📦 0.1.4
- skimage-normalize: fix typo in 0aca581e
- 📦 0.1.3
- all: fix pageId for sub-page derived images…
opencv-python 2ce6d97..a5fc295
Release: 46
- fix min supported numpy
- use numpy 1.19.3
- fix py version
- add arm64 matrix entries and update Dockerfile
- add test image
- test depends, more specific pyproject.toml
- add numpy version to setup.py
- add env markers for aarch64
- try with new images
- update README
- another fix for slashes
- fix slashes
- download Python 3.9 manually
- fix min numpy version
- fix py39 entry
- fix auditwheel patching
- fix line endings
- Merge branch 'feat/python39' of https://github.com/skvark/opencv-python into feat/python39
- fix pyproject.toml
- fix typo
- Python 3.9 preparations
sbb_binarization f0e0a75..30a3c98
Release: v0.0.5
- require h5py < 3, qurator-spk/sbb_textline_detection#50, tensorflow/tensorflow#44467
- 📦 v0.0.5
- Merge pull request #17 from sulzbals/ram-consumption
- Update README.md
- Update README.md
- Update README.md
- Update README.md
- Update README.md
- Update README.md
sbb_textline_detector 247d5f9..5d7c090
tesseract 5761880..e20ffdd
Release: 5.0.0-alpha
- Merge pull request #3153 from stweil/scale
workflow-configuration aa4fe96..a9a797e
- fix errexit in subshell
- add util fix-page-coords
- Merge branch 'master' of https://github.com/bertsky/workflow-configuration
- append to workspace.workflow log files (don't truncate)
v2018-11-18
typo in tag, please see https://github.com/OCR-D/ocrd_all/releases/v2020-11-18
v2020-10-29
cor-asv-ann f48b0e2..0ae6867
Release: v0.1.6
- evaluate: use output file group with JSON reports per page and totals
- lib.alignment: fix dist (use largest length as denominator)
- Merge branch 'fix-zip-input-files' into 'master'
dinglehopper f2367ac..6e47acd
- 📝 dinglehopper: Move screenshot higher
- 🐛 dinglehopper: Skip pages if there is no GT nor OCR (Fixes GH-34)
- 🐛 dinglehopper: Do try to get text when no TextEquivs exist
- Merge branch 'feat/text-extraction-levels'
ocrd_fileformat ed932e1..ded152d
Release: v0.1.1
- 📦 v0.1.1
- Merge pull request #20 from OCR-D/fix-overwrite
tesseract 11297c9..5761880
Release: 5.0.0-alpha
- Merge pull request #3141 from stweil/invert
workflow-configuration 1476350..aa4fe96
- prevent running pattern recipe without prereqs
- fix stop for implicit rule search for makefiles
v2020-10-22.1
v2020-10-22
core ac86b8e..6ca42f1
Release: v2.18.1
- 📦 v2.18.1
- 📝 changelog
- Merge pull request #631 from OCR-D/fix-logging-conf-override
- 📝 changelog
- [app][log] remove ocrd_workspace logger from default config
- [app][log] ocrd.workspace logger outcommented
- [app]log] all explicite configurations
- [app][doc] integrate loggers into default config
- [app][doc] cleared name
- 📝 changelog
- Merge pull request #628 from ulb-sachsen-anhalt/fix/cli-log-level
- 📦 v2.18.0
- 📝 changelog
- update spec to 3.9.1
- 📝 changelog
- Merge pull request #623 from bertsky/add-processor-help-docstring
- 📝 changelog
- Merge pull request #614 from OCR-D/find-files-generator
- 📦 v2.17.2
- 📝 changelog
- Merge remote-tracking branch 'stweil/numpy'
- 📦 v2.17.1
- 📝 changelog
- Merge branch 'master' of https://github.com/OCR-D/core
- 📝 changelog
- Merge pull request #618 from stweil/master
dinglehopper d706ef4..f2367ac
- 🐛 Fix OCR-D CLI for newest OCR-D
- ✨ dinglehopper: Show a progressbar on --progress
- 🧹 dinglehopper: Disable metrics in JSON report, too
- 📝 dinglehopper: Update README to use OCR-D's new and more readable -P option
- 🧹 dinglehopper: Move all normalization code to extracted_text.py
- Merge branch 'master' of https://github.com/qurator-spk/dinglehopper
- 📝 dinglehopper: Update screenshot to include a region id tooltip
- 📝 dinglehopper: Add detail about the text extraction and ExtractedText
- 📝 dinglehopper: Move developer info to README-DEV.md
- Merge branch 'feat/display-segment-id'
ocrd_anybaseocr a0999ec..cb82aad
Release: v1.0.2
- wip fix for #74
ocrd_calamari 04e950a..795826f
Release: v0.0.7
- Merge branch 'master' of https://github.com/OCR-D/ocrd_calamari
- 🧹 Do not install numpy, let the TF dependency do it
ocrd_cis a6f90ab..5ec0e34
Release: v0.1.5
- Merge pull request #75 from bertsky/fix-clip-mode-with-alpha
ocrd_olena cd7a359..9db846c
Release: v1.2.3
ocrd_pc_segmentation 5e8950f..7298475
Release: v0.1.3
- Really force updating of dependency versions on
make deps- Force regeneration of requirements.txt on
make deps- Bump version: 0.2.2 → 0.2.3
ocrd_typegroups_classifier f9bc14e..3d0004f
Release: v0.1.2
- 📦 v0.1.2
- Merge pull request #3 from OCR-D/getlogger-in-process
ocrd_wrap a79cd7d..1876b65
Release: v0.1.1
- 📦 0.1.2
- Merge branch 'master' of ssh://github.com/bertsky/ocrd_wrap
- skimage-normalize: expose black-point/white-point params
opencv-python c001042..2ce6d97
Release: 44
- Merge pull request #402 from stweil/gitignore
sbb_textline_detector cbc9838..247d5f9
- Merge pull request #45 from bertsky/readme-ocr-d-cli
- 🐛 sbb_textline_detector: Re-base Border coords too
- 🐛 sbb_textline_detector: Filter cropped images (OCR-D)
- Merge branch 'fix/image_from_page'
- 🐛 sbb_textline_detector: Work around incompatible numpy version installed by pip
tesseract b19e3ee..11297c9
Release: 5.0.0-alpha
- Merge pull request #3130 from robinwatts/pushback15
- Merge pull request #2994 from robinwatts/pushback11
- Merge pull request #3119 from Shatur95/modernize-cmake-config
- Merge pull request #3120 from amitdo/legacy
- Merge pull request #3116 from Shatur95/older-install-syntax
- Merge pull request #3114 from stweil/globaloc
- Merge pull request #3113 from stweil/pango
- Merge pull request #3112 from Shatur95/fix-cmake-install-configs
v2020-10-13
core 56c4aa6..ac86b8e (rewind)
Release: v2.17.0
📦 v2.18.0
📝 changelog
update spec to 3.9.1
📝 changelog
Merge pull request #623 from bertsky/add-processor-help-docstring
📝 changelog
Merge pull request #614 from OCR-D/find-files-generator
📦 v2.17.2
📝 changelog
Merge remote-tracking branch 'stweil/numpy'
📦 v2.17.1
📝 changelog
Merge branch 'master' of https://github.com/OCR-D/core
📝 changelog
Merge pull request #618 from stweil/master
ocrd_pc_segmentation 77aafcb..5e8950f (rewind)
Release: v0.1.3
Bump version: 0.2.2 → 0.2.3
opencv-python 2ce6d97..c001042 (rewind)
Release: 44
Merge pull request #402 from stweil/gitignore
tesseract ec01b51..b19e3ee (rewind)
Release: 5.0.0-alpha
Merge pull request #3119 from Shatur95/modernize-cmake-config
Merge pull request #3120 from amitdo/legacy
Merge pull request #3116 from Shatur95/older-install-syntax
Merge pull request #3114 from stweil/globaloc
Merge pull request #3113 from stweil/pango
Merge pull request #3112 from Shatur95/fix-cmake-install-configs
v2020-10-02
- Fix "core not updated in sub venv" issue #195
dinglehopper 7085ee0..d706ef4
- 📝 Document CER/WER and the format detection (Fixes GH-26)
- 💩 dinglehopper: Fix OCR-D CLI test by working around ocrd_cli_wrap_processor() check for arguments
ocrd_calamari bb9b1ab..04e950a
Release: v0.0.7
- Merge pull request #45 from OCR-D/getlogger
- 🚧 README: Mention METS
- 📝 Let intro mention ocrd_calamari + PAGE XML, link to OCR-D
ocrd_cis fb6cb47..a6f90ab
Release: v0.1.4
- Merge pull request #71 from bertsky/fix-getlogger-common
ocrd_fileformat 584240b..ed932e1 (rewind)
Release: v0.1.0
< respect --overwrite, fix #16
ocrd_ocropy 573d521..a6e556e
Release: v0.0.3
- 📝 Stres in README that project is obsolete
ocrd_olena 54792be..cd7a359
Release: v1.2.2
- Merge pull request #72 from bertsky/fix-typo-raise
ocrd_pagetopdf 58df37a..6047bd4
Release: v1.0.0
ocrd_pc_segmentation 3e69bff..5e8950f
Release: v0.1.3
- Update pixel classifier dependency
- .gitignore: Ignore generated requirements.txt
ocrd_repair_inconsistencies 8182063..c898d6c
ocrd_tesserocr 27fe2be..24b7ced
Release: v0.9.5
- Update CHANGELOG.md
- Merge pull request #156 from OCR-D/getlogger