A small script for quick and dirty image to pdf conversion with ocr.

bookbinder.sh

#!/bin/bash
#
# Bookbinder by Joel Bergroth
# A small script for quick and dirty image to pdf conversion with ocr. For example if your scanner produces multipaged tiffs and you want a small ocr'd pdf for reading on your tablet. 
#
# Requirements: imagemagick, tesseract, scantailor, pdfbeads
#
#
# Simple usage: sh bookbinder [/path/to/folder]
#
# Works with black and white multipage tiffs stored in a folder.
# Example: sh bookbinder exampledit
# Example: sh bookbinder /home/username/exampledir
# Example: sh bookbinder exampledir swe
# 
# Custom usage: sh bookbinder [/path/to/folder] [lang] [density] [color]
# Example: sh bookbinder exampledir swe 300 color

if [ -z "$1" ]; then
  echo "No input directory."
  exit 1
fi

if [ -z "$2" ]; then
  language=eng
else
  language=$2
fi

if [ -z "$3" ]; then
  density=""
else
  density="-density $3"
fi

if [ -z "$4" ]; then
  splitformat=tif
  colormode=""
else
  splitformat=png
  colormode="--color-mode=color_grayscale"
fi

cd $1
mkdir split
echo "Splitting multipage image files."
convert $density * split/%05d.$splitformat
mkdir pages
echo "Running scans through scantailor."
scantailor-cli $colormode split/* pages
echo "Running OCR with tesseract using language: $language"
cd pages
for img in *.tif; do
  filename=${img%%.*}
  tesseract $img $filename -l $language hocr 2>> ../tesseract-log.txt
  echo " Processed $img"
done
echo "Binding book using pdfbeads."
pdfbeads --bg-compression JPG --output ../../out.pdf *.tif
exit 0