This script will convert a pdf with double column to a pdf with one column. Some conference papers has double column and its hard to detect by OCR.

ChopTwoColumnPDF.sh

#!/bin/bash

# Converts a double-column PDF to a single-column PDF by extracting
# each column separately and then interleaving the pages.
#
# Usage:
#   ./2one.sh <input.pdf> <output.pdf>
#
# Arguments:
#   input.pdf:  The path to the input PDF file.
#   output.pdf: The path to the output PDF file.

# Check if correct arguments provided
if [ "$#" -ne 2 ]; then
  echo "Usage: $0 <input_pdf> <output_pdf>"
  exit 1
fi

INPUT_PDF="$1"
OUTPUT_PDF="$2"
CURRENT_DIR=$(pwd)

# If input doesn't contain path, assume current directory
if [[ "$INPUT_PDF" != */* ]]; then
  INPUT_PDF="$CURRENT_DIR/$INPUT_PDF"
fi

# If output doesn't contain path, assume current directory
if [[ "$OUTPUT_PDF" != */* ]]; then
  OUTPUT_PDF="$CURRENT_DIR/$OUTPUT_PDF"
fi

LEFT_COLUMN="$CURRENT_DIR/left_column_temp.pdf"
RIGHT_COLUMN="$CURRENT_DIR/right_column_temp.pdf"

# Get page size in points
echo "Getting page dimensions..."
page_width_pts=$(pdfinfo "$INPUT_PDF" | grep "Page size" | awk '{print $3}')
half_width_pts=$(echo "scale=0; $page_width_pts / 2" | bc)
echo "Page width in pts: $page_width_pts"
echo "Half width in pts: $half_width_pts"
# Extract left and right columns
echo "Extracting left column..."
pdfjam --papersize "{${half_width_pts}pt,22cm}" --trim "0pt 0pt ${half_width_pts}pt 0pt" --clip true "$INPUT_PDF" --outfile "$LEFT_COLUMN"

echo "Extracting right column..."
pdfjam --papersize "{${half_width_pts}pt,22cm}" --trim "${half_width_pts}pt 0pt 0pt 0pt" --clip true "$INPUT_PDF" --outfile "$RIGHT_COLUMN"

# Count pages in each PDF
pages_left=$(pdfinfo "$LEFT_COLUMN" | grep Pages | awk '{print $2}')
pages_right=$(pdfinfo "$RIGHT_COLUMN" | grep Pages | awk '{print $2}')

# Create shuffle pattern
pattern=""
for ((i=1; i<=pages_left; i++)); do
  pattern+="A$i B$i "
done

# Run pdftk with generated pattern
# The issue was here - can't use both shuffle and cat in the same command
pdftk A="$LEFT_COLUMN" B="$RIGHT_COLUMN" shuffle $pattern output "$OUTPUT_PDF"

# Clean up temporary files
rm "$LEFT_COLUMN" "$RIGHT_COLUMN"

echo "Created $OUTPUT_PDF with alternating pages from left and right columns of $INPUT_PDF"