Script to OCR many PDF files using tessarect and imagemagick

ocr.sh

#!/bin/bash

set -e

analyze() {
	BASE="$1"
	echo "Converting 'input/$BASE.pdf' to 'output/$BASE.tiff'"
	convert -density 300 "input/$BASE.pdf" "output/$BASE.tiff" 2>&1
	
	echo "OCR of 'output/$BASE.pdf' to 'output/${BASE}_deu.pdf'"
	tesseract "output/$BASE.tiff" "output/${BASE}_deu" -l deu --dpi 300 pdf 2>&1

	echo "Removing 'output/$BASE.tiff'"
	rm "output/$BASE.tiff"
}

mkdir -p output

i=0
parallel=4
for PDF in input/*.pdf; do
	BASE="${PDF%.*}"
	BASE="${BASE##*/}"

	echo "Starting process for 'output/$BASE.pdf' in the background"
	analyze "$BASE" 2>&1 > "output/$BASE.log" &
	if [ $(( i++ % parallel )) == 0 ]; then
		echo "Waiting on $parallel process to complete"
		wait
	fi
done

echo "Waiting for completion"
wait