OCR de documentos con Imagemagick y Tesseract

ocr.sh

#!/bin/sh

STARTPAGE=1 # set to pagenumber of the first page of PDF you wish to convert
ENDPAGE=11 # set to pagenumber of the last page of PDF you wish to convert
SOURCE=source.pdf # set to the file name of the PDF
OUTPUT=destination.txt # set to the final output file
RESOLUTION=75 # set to the resolution the scanner used (for B/W and good scans 75 suffice)

touch $OUTPUT
for i in `seq $STARTPAGE $ENDPAGE`; do
    echo extracting page $i
    convert -monochrome -density $RESOLUTION $SOURCE\[$(($i - 1 ))\] /tmp/page.tif
    echo processing page $i
    tesseract -l spa -psm 3 /tmp/page.tif /tmp/tempoutput
    cat /tmp/tempoutput.txt >> $OUTPUT
done