Extract images from pdf but, surprise! The images that contain alpha channels are extracted as two images. The source and the mask, as they are stored inside pdf file. This script recombine them into an one png file with given transparency

pdfimage

#/usr/bin/bash
# extract images from pdf at first-hand
#prefix=pict
#echo extract images from "$1"
#pdfimages -j "$1" $prefix
# IMAGES ARE SAVED SECVENTIALLY AS IMAGE AND THE NEXT IS A MASK IMAGE !!!!
declare -a files=($(ls *.ppm *.pbm))
mask=
image=
for (( i = 0; i < ${#files[*]}; i = i + 2 ))
do
  image=${files[$i]}
  mask=${files[$i+1]}
  echo converting "${image%.*}".png
  convert $image $mask -compose CopyOpacity -composite "${image%.*}".png
  echo removing image and it\'s mask
  rm "$image" "$mask"
done

echo done!