From 2ebf3c0e005ce069e4a2ff05c2ca76d050700a54 Mon Sep 17 00:00:00 2001
From: Konstantin Baierer <unixprog@gmail.com>
Date: Thu, 26 Jul 2018 19:09:07 +0200
Subject: [PATCH] .

---
 .gitignore                   |  2 ++
 Makefile                     | 11 ++++++++++
 ocrd-tool.json               |  1 +
 ocrd_calamari/ocr.py         | 39 ++++++++++++++++++++++++++++++++++++
 ocrd_calamari/ocrd-tool.json | 25 +++++++++++++++++++++++
 5 files changed, 78 insertions(+)
 create mode 100644 Makefile
 create mode 120000 ocrd-tool.json
 create mode 100644 ocrd_calamari/ocr.py
 create mode 100644 ocrd_calamari/ocrd-tool.json

diff --git a/.gitignore b/.gitignore
index 894a44c..a67a4c3 100644
--- a/.gitignore
+++ b/.gitignore
@@ -102,3 +102,5 @@ venv.bak/
 
 # mypy
 .mypy_cache/
+/calamari
+/calamari_models
diff --git a/Makefile b/Makefile
new file mode 100644
index 0000000..2f88071
--- /dev/null
+++ b/Makefile
@@ -0,0 +1,11 @@
+GIT_CLONE = git clone --depth 1
+calamari:
+	$(GIT_CLONE) https://github.com/chwick/calamari
+
+calamari_models:
+	$(GIT_CLONE) https://github.com/chwick/calamari_models
+
+calamari/build: calamari calamari_models
+	cd calamari &&\
+		pip install -r requirements.txt ;\
+		python setup.py install
diff --git a/ocrd-tool.json b/ocrd-tool.json
new file mode 120000
index 0000000..4a7b986
--- /dev/null
+++ b/ocrd-tool.json
@@ -0,0 +1 @@
+ocrd_calamari/ocrd-tool.json
\ No newline at end of file
diff --git a/ocrd_calamari/ocr.py b/ocrd_calamari/ocr.py
new file mode 100644
index 0000000..2dd6038
--- /dev/null
+++ b/ocrd_calamari/ocr.py
@@ -0,0 +1,39 @@
+from __future__ import absolute_import
+from calamari_ocr.scripts.predict import run
+
+log = getLogger('processor.KrakenOcr')
+
+class KrakenOcr(Processor):
+
+    def __init__(self, *args, **kwargs):
+        kwargs['ocrd_tool'] = OCRD_TOOL['tools']['ocrd-calamari-ocr']
+        super(KrakenOcr, self).__init__(*args, **kwargs)
+
+    def process(self):
+        """
+        Performs the binarization.
+        """
+        for (n, input_file) in enumerate(self.input_files):
+            log.info("INPUT FILE %i / %s", n, input_file)
+            pcgts = ocrd_page.from_file(self.workspace.download_file(input_file))
+            image_url = pcgts.get_Page().imageFilename
+            log.info("pcgts %s", pcgts)
+            for region in pcgts.get_Page().get_TextRegion():
+                textlines = region.get_TextLine()
+                log.info("About to binarize %i lines of region '%s'", len(textlines), region.id)
+                for (line_no, line) in enumerate(textlines):
+                    log.debug("Binarizing line '%s' in region '%s'", line_no, region.id)
+                    image = self.workspace.resolve_image_as_pil(image_url, polygon_from_points(line.get_Coords().points))
+                    print(dir(kraken.binarization))
+                    bin_image = kraken.binarization.nlbin(image)
+                    bin_image_bytes = io.BytesIO()
+                    bin_image.save(bin_image_bytes, format='PNG')
+                    ID = concat_padded(self.output_file_grp, n)
+                    self.add_output_file(
+                        ID=ID,
+                        file_grp=self.output_file_grp,
+                        basename="%s.bin.png" % ID,
+                        mimetype='image/png',
+                        content=bin_image_bytes.getvalue()
+                    )
+
diff --git a/ocrd_calamari/ocrd-tool.json b/ocrd_calamari/ocrd-tool.json
new file mode 100644
index 0000000..6c0b0ad
--- /dev/null
+++ b/ocrd_calamari/ocrd-tool.json
@@ -0,0 +1,25 @@
+{
+  "git_url": "https://github.com/OCR-D/ocrd_calamari",
+  "version": "0.0.1",
+  "tools": {
+    "ocrd-calamari-ocr": {
+      "executable": "ocrd-calamari-ocr",
+      "categories": [
+        "Text recognition and optimization"
+      ],
+      "steps": [
+        "recognition/text-recognition"
+      ],
+      "description": "Recognize lines with kraken",
+      "parameters": {
+        "checkpoint": {"type": "string", "format": "file", "cacheable": true},
+        "processes": {"type": "number", "default": 1},
+        "batch_size": {"type": "number", "default": 1},
+        "voter": {"type": "string", "default": "confidence_voter_default_ctc"},
+        "extended_prediction_data_format": {"type": "string", "default": "json"},
+        "XXX output_dir": "TODO",
+        "XXX extended_prediction_data": "TODO"
+      }
+    }
+  }
+}