📦 0.1.17

CHANGELOG.md

@@ -4,6 +4,12 @@ Versioned according to [Semantic Versioning](http://semver.org/).
 
 ## Unreleased
 
+## [0.1.17] - 2022-03-12
+
+Changed:
+
+ * project: use true alpha shape instead of convex hull
+
 ## [0.1.16] - 2022-02-21
 
 Fixed:

README.md

@@ -26,7 +26,7 @@ Contains processors for various tasks:
     - [ocrd-segment-from-coco](ocrd_segment/import_coco_segmentation.py) (for [MS-COCO](https://cocodataset.org/) annotation)
   - post-processing or repairing layout segmentations:
     - [ocrd-segment-repair](ocrd_segment/repair.py) (validity and consistency of coordinates, reducing overlaps/redundancy between neighbours, shrinking regions to the alpha shape of their lines)
-    - [ocrd-segment-project](ocrd_segment/project.py) (remake segment coordinates into the convex hull of their constituents)
+    - [ocrd-segment-project](ocrd_segment/project.py) (remake segment coordinates into the concave hull / alpha shape of their constituents)
     - [ocrd-segment-replace-original](ocrd_segment/replace_original.py) (rebase all segments on cropped+deskewed border frame as new full page)
     - [ocrd-segment-replace-page](ocrd_segment/replace_page.py) (2 input fileGrps; overwrite segmentation below page of first fileGrp by all segments of second fileGrp, rebasing all coordinates; "inverse" of `replace-original`)
   - comparing different layout segmentations:

ocrd_segment/ocrd-tool.json

@@ -1,5 +1,5 @@
 {
-  "version": "0.1.16",
+  "version": "0.1.17",
   "git_url": "https://github.com/OCR-D/ocrd_segment",
   "tools": {
     "ocrd-segment-repair": {

ocrd_segment/repair.py

@@ -88,7 +88,7 @@ def process(self):
         
         Furthermore, if ``sanitize``, then for each text region, update
         the coordinates to become the minimal convex hull of its constituent
-        text lines.
+        text lines. (But consider running ocrd-segment-project instead.)
         
         Finally, produce new output files by serialising the resulting hierarchy.
         """
@@ -254,6 +254,7 @@ def plausibilize_page(self, page, page_id):
 
     def sanitize_page(self, page, page_id):
         """Shrink each region outline to become the minimal convex hull of its constituent textlines."""
+        # FIXME: should probably be removed in favour of ocrd-segment-project entirely
         LOG = getLogger('processor.RepairSegmentation')
         regions = page.get_AllRegions(classes=['Text'])
         page_image, page_coords, _ = self.workspace.image_from_page(