print_materials/ocr/draw_overlays.ipynb

+%% Cell type:code id: tags:
+
+``` python
+from lxml import etree as et
+from pathlib import Path
+from PIL import Image, ImageDraw
+
+ns = {
+    'tei': 'http://www.tei-c.org/ns/1.0',
+    'xml': 'http://www.w3.org/XML/1998/namespace'
+}
+
+tei_path = Path('anno-event-2023_tei.xml')
+```
+
+%% Cell type:code id: tags:
+
+``` python
+root = et.parse(tei_path).getroot()
+facs = root.find('.//tei:facsimile[@xml:id="facs_5"]', namespaces=ns)
+
+root, facs
+```
+
+%% Cell type:code id: tags:
+
+``` python
+colors = {
+    'TextRegion': {
+        'fill': '#d9e1f2',
+        'outline': None
+    },
+    'Image': {
+        'fill': '#fce4d6',
+        'outline': None
+    },
+    'UnknownRegion': {
+        'fill': '#fff2cc',
+        'outline': None
+    },
+    'Line': {
+        'fill': '#f8cbad',
+        'outline': 'grey'
+    },
+    'Word': {
+        'fill': '#bdd7ee',
+        'outline': 'grey'
+    }
+}
+
+def draw_annotations(image_path, rends):
+    img = Image.open(image_path)
+    width, height = img.size
+    converted = img.convert("RGBA")
+    image_data = converted.tobytes()
+    img = Image.frombytes('RGBA', (width, height), image_data)
+    for facs in root.findall('.//tei:zone', namespaces=ns):
+        if facs.attrib['rendition'] in rends:
+            rect = ImageDraw.Draw(img)
+            shape = [
+                (int(facs.attrib['ulx']), int(facs.attrib['uly'])),
+                (int(facs.attrib['lrx']), int(facs.attrib['lry'])),
+            ]
+            fill = colors[facs.attrib['rendition']]['fill']
+            outline = colors[facs.attrib['rendition']]['outline']
+            rect = rect.rectangle(shape, fill=fill, outline=outline)
+
+
+
+    img.save(image_path)
+```
+
+%% Cell type:code id: tags:
+
+``` python
+
+draw_annotations(Path('../5_region_segmentation.png'), ['TextRegion', 'Image', 'UnknownRegion'])
+draw_annotations(Path('../6_line_detection.png'), ['Line'])
+draw_annotations(Path('../7_word_detection.png'), ['Word'])
+```

print_materials/ocr/requirements.txt

+lxml==4.9.2
+Pillow==10.1.0
\ No newline at end of file