A shel script to extracts images from the passed PDF file using pdfimages, and further process them.

extract-image-from-page-pdf.sh

#!/bin/bash
## By Ahmad Gharbeia, January 2016. Licensed under GPL version 3.0
## Extracts images from the passed PDF file using pdfimages.
## Additionally, converts PNM files (PBM and PPM) to PNG and optimises the resulting file using pnmtopng and pngcrush respectively.
## JPEG images resulting from the extraction of DCT images from the source PDF, are left unchanged.

dir=${1:-'.'}

shopt -s nullglob

tempdir=$(mktemp -d --tmpdir "$(basename $0).XXXXXXXXXX")

for pdf in "$dir"/*.pdf
do
	if [ $(pdfimages -list "$pdf" | grep -c image) -eq 1 ]; then
		echo "$pdf: "
		pdfimages -j "$pdf" "${pdf%.pdf}"
	fi
	for pnm in "${pdf%.pdf}"-???.p{b,p}m
	do
		echo  "$pnm" "-> ${pnm:0: -4}".png
		temppng=$(mktemp --tmpdir=$tempdir XXXXXXXXXX)
##code is needed for conditionally rotating extracted images.
#		pnmrotate -noantialias 90 "$pnm" | pnmtopng > $temppng
		pnmtopng "$pnm" > $temppng
		rm "$pnm"
		pngcrush -fix -l 1 -q -z 9 $temppng "${pnm:0: -4}".png
		rm $temppng
	done
done
rm -r $tempdir