Extract OCR contents in PDF-documents as plain text (.txt)

pdf-ocr-to-txt.sh

# Extract OCR contents in PDF-documents as plain text (.txt)
# Kristian Risager Larsen, 2024-07
#
# Setup:
# You need to install GhostsScript and Tesseract
# brew install tesseract tesseract-lang ghostscript
#
# Notes:
# The "-l dan" parameter tells Tesseract to expect Danish text

for filename in *.pdf; do
  gs -dNOPAUSE -sDEVICE=pngalpha -r300 -sOutputFile="${filename%.pdf}%03d.png" "$filename" -c quit  
done

for filename in *.png; do
    tesseract "$filename" "${filename%.png}.txt" -l dan
done