pdf_to_ocr_text.sh

#!/bin/bash 

mkdir -p images/

if [[ $# -ne 1 ]]; then
  echo provide the pdf file as a single argument.
  exit 1
fi

f=$1

echo $f
o=$(echo $f | sed "s/.pdf//")

pdfimages $f images/$o

ls images/$o-*ppm > tmp

echo "debug_file /dev/null" > quiet.conf
tesseract tmp $o.ocr -l eng --dpi 200 pdf txt quiet.conf

rm tmp images/$o-*