Extracts images embedded in a PDF file to a TIFF image file

pdfscan2tiff.sh

#!/bin/sh
# Extracts images embedded in a PDF file (such PDFs created from document scanning) to a TIFF image file.

# SUSE Bash - https://documentation.suse.com/sles/12-SP4/html/SLES-all/cha-adm-shell.html
#  - General shell scripting - https://www.shellscript.sh/
# pdfimages manual - https://www.xpdfreader.com/pdfimages-man.html
#  - NOTE: use the "-list" option to retreive details (including resolution) for each image
#  - c.f https://stackoverflow.com/a/11699848
#  - see also https://graphicdesign.stackexchange.com/q/5670
# tiffcp manual - https://libtiff.gitlab.io/libtiff/tools/tiffcp.html

# For openSUSE Tumbleweed, package dependencies include the following:
#  - poppler-tools
#  - tiff

# As the tiff2pdf command has been removed from the tiff package,
#  consider "convert" in the ImageMagick package as an alternative.
#  - https://www.reddit.com/r/voidlinux/comments/197fy0n/?rdt=58586

if [ -z "$1" ] || [ ! -e "$1" ] ; then
  echo "Usage: `basename $0` <Input PDF File> [Output TIFF File]"
else
  INPUT_FILE="$1"
  OUTPUT_FILE=$([ -z "$2" ] && echo "${INPUT_FILE%.*}.tif" || echo "$2")
  TIFF_TEMP_DIR=`mktemp -d tiff-page-cache_XXXXXXX --tmpdir`
  TIFF_TEMP_FILE=$TIFF_TEMP_DIR/rotated-split.tif

  pdfimages -tiff "$INPUT_FILE" "$TIFF_TEMP_DIR/page"
  tiffcp -c zip $TIFF_TEMP_DIR/*.tif "$OUTPUT_FILE"
  
  # Optionally handle full-spread book scans
  if ! true ; then
    tiffcrop -R 90 -S 2:1 "$OUTPUT_FILE" "$TIFF_TEMP_FILE"
    mv -f "$TIFF_TEMP_FILE" "$OUTPUT_FILE"
  fi
  trap "rm -d -r '$TIFF_TEMP_DIR'" EXIT
fi