Don’t rotate pages when using pdfjoin
pdfjoin --paper a4paper --rotateoversize false <input.pdf>
Tagpdf
pdfsandwich: OCR PDFs containing images
pdfsandwich is a handy tool developed by Tobias Elze for OCR’ing (via tesseract) scanned documents. Recognized text is added as an background layer, making it possible to search and index scanned documents.
pdfsandwich -lang eng+deu scanned.pdf
PDF to JPG using ImageMagick’s convert
…while dealing with alpha/transparency:
convert -verbose -density 150 -trim <input>[PAGE-RANGE] -quality 100 -sharpen 0x1.0 -background white -alpha remove <output>
pdf2jpg.sh, usage: pdf2jpg.sh <input> [page-range], page range starting at 0.
#!/bin/bash
INPUT=$1
PAGES=$2
ME=`basename "$0"`
if [[ ! -f "${INPUT}" ]]
then
echo "Input not found"
echo "Usage: ${ME} <pdf> [[pages]]"
exit 1
fi
if [[ $(file --mime-type -b "${INPUT}") != "application/pdf" ]]
then
echo "Input not a PDF"
exit 1
fi
BASENAME="`basename "${INPUT}" .pdf`"
OUTPUT=$(mktemp -q -u "${BASENAME}.XXXXXXXXX")
convert -verbose -density 150 -trim "${INPUT}${PAGES}" -quality 100 -sharpen 0x1.0 -background white -alpha remove "${OUTPUT}-%03d.jpg"