From cd0c0a821361f5ee7c52ee60fb0ed5b758e53620 Mon Sep 17 00:00:00 2001
From: Nick White <git@njw.me.uk>
Date: Sun, 30 Oct 2011 12:29:45 +0000
Subject: Add ocr pdf script

---
 extras/mkocrpdf-cuneform.sh | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)
 create mode 100644 extras/mkocrpdf-cuneform.sh

(limited to 'extras/mkocrpdf-cuneform.sh')

diff --git a/extras/mkocrpdf-cuneform.sh b/extras/mkocrpdf-cuneform.sh
new file mode 100644
index 0000000..94e10eb
--- /dev/null
+++ b/extras/mkocrpdf-cuneform.sh
@@ -0,0 +1,21 @@
+#!/bin/sh
+#
+# Makes a pdf with embedded text as extracted by cuneiform
+#
+# Requires imagemagick, pdftk, hocr2pdf and cuneiform
+
+for i in `ls *png`
+do
+	a=`basename $i .png`
+	echo processing $a
+
+	convert $i $a.bmp
+	cuneiform -f hocr -o $a.html $a.bmp
+	rm -f $a.bmp
+
+	# hocr2pdf has a habit of segfaulting, so fall back to convert
+	hocr2pdf -i $i -o $a.pdf < $a.html || convert $i $a.pdf
+	rm -f $a.html
+done
+
+pdftk *pdf cat output book.pdf
-- 
cgit v1.2.3