OCR all files in a folder using Tesseract, ignoring existing files.

ocr.bash

#! /usr/bin/env bash

### ocr
#
# OCR all files in a folder using Tesseract, ignoring existing files.
#

## Functions

usage() {
  echo "Usage: ocr <source-directory> <destination-directory>"
  exit 1
}

if ! [ -x "$(command -v tesseract)" ]; then
  echo -e 'Tesseract is required to use this script.\n\nFor more information, visit: https://github.com/tesseract-ocr/tesseract' >&2
  exit 1
fi

## Arguments

if (( $# != 2 ))
then
  usage
fi

source="$1"
dest="$2"

## Main

mkdir -p "$dest"

shopt -s nullglob
shopt -s nocaseglob

for f in "$source"/*.{png,jpg,jpeg}; do
  filename=`basename "$f"`

  if [ ! -f "$dest/$filename.txt" ]; then
    echo "PROCESSING $f"
    tesseract "$f" "$dest/$filename"
  fi
done

shopt -u nocaseglob
shopt -u nullglob

if I do bash ocr.bash I get Usage: ocr <source-directory> <destination-directory>
what do I do if my goal is to convert all the files in a folder but retain their original names?
something like: ``tesseract *.jpg *.txt` (pseudocode)