📦 0.1.13

CHANGELOG.md

@@ -4,19 +4,33 @@ Versioned according to [Semantic Versioning](http://semver.org/).
 
 ## Unreleased
 
-## [0.1.12]
+## [0.1.13] - 2021-12-11
+
+Fixed:
+
+ * evaluate: multi-matching (without pycocotools)
+
+Changed:
+
+ * evaluate: improved report format (hierarchy and names)
+
+Added:
+
+ * evaluate: over-/undersegmentation metrics, pixel-wise metrics
+
+## [0.1.12] - 2021-12-03
 
 Changed:
 
  * evaluate: basic IoU matching, Pr/Rc and mAP/mAR stats via pycocotools
 
-## [0.1.11]
+## [0.1.11] - 2021-03-23
 
 Fixed:
 
  * extract-pages: `Border` has no `id`
 
-## [0.1.10]
+## [0.1.10] - 2021-02-26
 
 Fixed:
 
@@ -27,13 +41,13 @@ Changed:
  * extract-pages: add `feature_filter` param
  * extract-pages: add `order` choice for `plot_segmasks`
 
-## [0.1.9]
+## [0.1.9] - 2021-02-10
 
 Changed:
 
  * extract-regions/lines/words/glyphs: add `feature_filter` param
 
-## [0.1.8]
+## [0.1.8] - 2021-02-07
 
 Fixed:
 
@@ -46,22 +60,22 @@ Changed:
  * extract-pages: expose `colordict` parameter (w/ same default)
  * extract-pages: multi-level mask output via `plot_segmasks`
 
-## [0.1.7]
+## [0.1.7] - 2021-01-07
 
 Fixed:
 
  * repair: also ensure polygons have at least 3 points
  * replace-page: allow non-PAGE input files, too
 
-## [0.1.6]
+## [0.1.6] - 2020-11-25
 
 Fixed:
 
  * repair: also fix negative coords, also on page level
  * replace-original: also remove page border/@orientation
  * replace-original: add new original as derived image, too
 
-## [0.1.5]
+## [0.1.5] - 2020-11-04
 
 Fixed:
 
@@ -72,7 +86,7 @@ Changed:
  * replace-original: delegate to `repair.ensure_consistent`
  * replace-page: new CLI (inverse or replace-original)
 
-## [0.1.4]
+## [0.1.4] - 2020-11-04
 
 Changed:
 
@@ -118,6 +132,7 @@ Changed:
   * further improve README
 
 <!-- link-labels -->
+[0.1.13]: ../../compare/v0.1.12...v0.1.13
 [0.1.12]: ../../compare/v0.1.11...v0.1.12
 [0.1.11]: ../../compare/v0.1.10...v0.1.11
 [0.1.10]: ../../compare/v0.1.9...v0.1.10

ocrd_segment/ocrd-tool.json

@@ -1,5 +1,5 @@
 {
-  "version": "0.1.12",
+  "version": "0.1.13",
   "git_url": "https://github.com/OCR-D/ocrd_segment",
   "tools": {
     "ocrd-segment-repair": {

setup.py

@@ -13,18 +13,6 @@
     - ocrd-segment-replace-original
     - ocrd-segment-replace-page
     - ocrd-segment-evaluate
-<<<<<<< HEAD
-=======
-    - page-segment-evaluate
-    - ocrd-segment-extract-address
-    - ocrd-segment-extract-formdata
-    - ocrd-segment-classify-address-text
-    - ocrd-segment-classify-address-layout
-    - ocrd-segment-classify-formdata-dummy
-    - ocrd-segment-classify-formdata-text
-    - ocrd-segment-classify-formdata-layout
-    - ocrd-segment-postcorrect-formdata
->>>>>>> e000d13... evaluate: refactor and expose standalone CLI
 """
 import codecs