Is there a simple way to identify if a PDF is scanned?

Solution 1:

1. Put all the .pdf files in one folder.
2. No .txt file in that folder.
3. In terminal change directory to that folder with cd <path to dir>
4. Make one more directory for non scanned files. Example:

mkdir ./x 
for file in *.pdf; do
    if [ $(pdftotext "$file")"x" == "x" ] ; then mv "$file" ./x; fi
rm *.txt
done

All the pdf scanned files will remain in the folder and other files will move to another folder.

Solution 2:

Shellscript

• If a pdf file contains an image (inserted in a document alongside text or as whole pages, 'scanned pdf'), the file often (maybe always) contains the string /Image/.

• In the same way you can search for the string /Text to tell if a pdf file contains text (not scanned).

I made the shellscript pdf-text-or-image, and it might work in most cases with your files. The shellscript looks for the text strings /Image/ and /Text in the pdf files.

#!/bin/bash

echo "shellscript $0"
ls --color --group-directories-first
read -p "Is it OK to use this shellscript in this directory? (y/N) " ans
if [ "$ans" != "y" ]
then
 exit
fi

mkdir -p scanned
mkdir -p text
mkdir -p "s-and-t"

for file in *.pdf
do
 grep -aq '/Image/' "$file"
 if [ $? -eq 0 ]
 then
  image=true
 else
  image=false
 fi
 grep -aq '/Text' "$file"
 if [ $? -eq 0 ]
 then
  text=true
 else
  text=false
 fi


 if $image && $text
 then
  mv "$file" "s-and-t"
 elif $image
 then
  mv "$file" "scanned"
 elif $text
 then
  mv "$file" "text"
 else
  echo "$file undecided"
 fi
done

Make the shellscript executable,

chmod ugo+x pdf-text-or-image

Change directory to where you have the pdf files and run the shellscript.

Identified files are moved to the following subdirectories

• scanned
• text
• s-and-t (for documents with both [scanned?] images and text content)

Unidentified file objects, 'UFOs', remain in the current directory.

Test

I tested the shellscript with two of your files, AR-G1002.pdf and AR-G1003.pdf, and with some own pdf files (that I have created using Libre Office Impress).

$ ./pdf-text-or-image
shellscript ./pdf-text-or-image
s-and-t                                 mkUSB-quick-start-manual-11.pdf    mkUSB-quick-start-manual-nox-11.pdf
scanned                                 mkUSB-quick-start-manual-12-0.pdf  mkUSB-quick-start-manual-nox.pdf
text                                    mkUSB-quick-start-manual-12.pdf    mkUSB-quick-start-manual.pdf
AR-G1002.pdf                            mkUSB-quick-start-manual-74.pdf    OBI-quick-start-manual.pdf
AR-G1003.pdf                            mkUSB-quick-start-manual-75.pdf    oem.pdf
DescriptionoftheOneButtonInstaller.pdf  mkUSB-quick-start-manual-8.pdf     pdf-text-or-image
GrowIt.pdf                              mkUSB-quick-start-manual-9.pdf     pdf-text-or-image0
list-files.pdf                          mkUSB-quick-start-manual-bas.pdf   README.pdf
Is it OK to use this shellscript in this directory? (y/N) y

$ ls -1 *
pdf-text-or-image
pdf-text-or-image0

s-and-t:
DescriptionoftheOneButtonInstaller.pdf
GrowIt.pdf
mkUSB-quick-start-manual-11.pdf
mkUSB-quick-start-manual-12-0.pdf
mkUSB-quick-start-manual-12.pdf
mkUSB-quick-start-manual-8.pdf
mkUSB-quick-start-manual-9.pdf
mkUSB-quick-start-manual.pdf
OBI-quick-start-manual.pdf
README.pdf

scanned:
AR-G1002.pdf

text:
AR-G1003.pdf
list-files.pdf
mkUSB-quick-start-manual-74.pdf
mkUSB-quick-start-manual-75.pdf
mkUSB-quick-start-manual-bas.pdf
mkUSB-quick-start-manual-nox-11.pdf
mkUSB-quick-start-manual-nox.pdf
oem.pdf

Let us hope that

• there are no UFOs in your set of files
• the sorting is correct concerning text versus scanned/images

Solution 3:

If this is more about actually detecting if PDF was created by scanning rather than pdf has images instead of text then you might need to dig into the metadata of the file, not just content.

In general, for the files I could find on my computer and your test files, following is true:

• Scanned files have less than 1000chars/page vs. non scanned ones who always have more than 1000chars/page
• Multiple independent scanned files had "Canon" listed as the PDF creator, probably referencing Canon scanner software
• PDFs with "Microsoft Word" as creator are likely to not be scanned, as they are word exports. But someone could scan to word, then export to PDF - some people have very strange workflow.

I'm using Windows at the moment, so I used node.js for the following example:

const fs = require("mz/fs");
const pdf_parse = require("pdf-parse");
const path = require("path");


const SHOW_SCANNED_ONES = process.argv.indexOf("scanned") != -1;

const DEBUG = process.argv.indexOf("debug") != -1;
const STRICT = process.argv.indexOf("strict") != -1;

const debug = DEBUG ? console.error : () => { };

(async () => {
    const pdfs = (await fs.readdir(".")).filter((fname) => { return fname.endsWith(".pdf") });

    for (let i = 0, l = pdfs.length; i < l; ++i) {
        const pdffilename = pdfs[i];
        try {
            debug("\n\nFILE: ", pdffilename);
            const buffer = await fs.readFile(pdffilename);
            const data = await pdf_parse(buffer);

            if (!data.info)
                data.indo = {};
            if (!data.metadata) {
                data.metadata = {
                    _metadata: {}
                };
            }


            // PDF info
            debug(data.info);
            // PDF metadata
            debug(data.metadata);
            // text length
            const textLen = data.text ? data.text.length : 0;
            const textPerPage = textLen / (data.numpages);
            debug("Text length: ", textLen);
            debug("Chars per page: ", textLen / data.numpages);
            // PDF.js version
            // check https://mozilla.github.io/pdf.js/getting_started/
            debug(data.version);

            if (evalScanned(data, textLen, textPerPage) == SHOW_SCANNED_ONES) {
                console.log(path.resolve(".", pdffilename));
            }
        }
        catch (e) {
            if (strict && !debug) {
                console.error("Failed to evaluate " + item);
            }
            {
                debug("Failed to evaluate " + item);
                debug(e.stack);
            }
            if (strict) {
                process.exit(1);
            }
        }
    }
})();
const IS_CREATOR_CANON = /canon/i;
const IS_CREATOR_MS_WORD = /microsoft.*?word/i;
// just defined for better clarity or return values
const IS_SCANNED = true;
const IS_NOT_SCANNED = false;
function evalScanned(pdfdata, textLen, textPerPage) {
    if (textPerPage < 300 && pdfdata.numpages>1) {
        // really low number, definitelly not text pdf
        return IS_SCANNED;
    }
    // definitelly has enough text
    // might be scanned but OCRed
    // we return this if no 
    // suspition of scanning is found
    let implicitAssumption = textPerPage > 1000 ? IS_NOT_SCANNED : IS_SCANNED;
    if (IS_CREATOR_CANON.test(pdfdata.info.Creator)) {
        // this is always scanned, canon is brand name
        return IS_SCANNED;
    }
    return implicitAssumption;
}

To run it, you need to have Node.js installed (should be a single command) and you also need to call:

npm install mz pdf-parse

Usage:

node howYouNamedIt.js [scanned] [debug] [strict]

 - scanned show PDFs thought to be scanned (otherwise shows not scanned)
 - debug shows the debug info such as metadata and error stack traces
 - strict kills the program on first error

This example is not considered finished solution, but with the debug flag, you get some insight into meta information of a file:

FILE:  BR-L1411-3-scanned.pdf
{ PDFFormatVersion: '1.3',
  IsAcroFormPresent: false,
  IsXFAPresent: false,
  Creator: 'Canon ',
  Producer: ' ',
  CreationDate: 'D:20131212150500-03\'00\'',
  ModDate: 'D:20140709104225-03\'00\'' }
Metadata {
  _metadata:
   { 'xmp:createdate': '2013-12-12T15:05-03:00',
     'xmp:creatortool': 'Canon',
     'xmp:modifydate': '2014-07-09T10:42:25-03:00',
     'xmp:metadatadate': '2014-07-09T10:42:25-03:00',
     'pdf:producer': '',
     'xmpmm:documentid': 'uuid:79a14710-88e2-4849-96b1-512e89ee8dab',
     'xmpmm:instanceid': 'uuid:1d2b2106-a13f-48c6-8bca-6795aa955ad1',
     'dc:format': 'application/pdf' } }
Text length:  772
Chars per page:  2
1.10.100
D:\web\so-odpovedi\pdf\BR-L1411-3-scanned.pdf

The naive function that I wrote has 100% success on the documents that I could find on my computer (including your samples). I named the files based on what their status was before running the program, to make it possible to see if results are correct.

D:\xxxx\pdf>node detect_scanned.js scanned
D:\xxxx\pdf\AR-G1002-scanned.pdf
D:\xxxx\pdf\AR-G1002_scanned.pdf
D:\xxxx\pdf\BR-L1411-3-scanned.pdf
D:\xxxx\pdf\WHO_TRS_696-scanned.pdf

D:\xxxx\pdf>node detect_scanned.js
D:\xxxx\pdf\AR-G1003-not-scanned.pdf
D:\xxxx\pdf\ASEE_-_thermoelectric_paper_-_final-not-scanned.pdf
D:\xxxx\pdf\MULTIMODE ABSORBER-not-scanned.pdf
D:\xxxx\pdf\ReductionofOxideMineralsbyHydrogenPlasma-not-scanned.pdf

You can use the debug mode along with a tiny bit of programming to vastly improve your results. You can pass the output of the program to other programs, it will always have one full path per line.

Solution 4:

I created a script to detect whether a PDF was OCRd. The main idea: In OCRd PDFs is the text invisible.

Algorithm to test whether a given PDF (f1) was OCRd:

1. create a copy of f1 noted as f2
2. delete all text on f2
3. create images (PNG) for all (or just a few) pages for f1 and f2
4. f1 was OCRd if all the images of f1 and f2 are identical.

https://github.com/jfilter/pdf-scripts/blob/master/is_ocrd_pdf.sh

#!/usr/bin/env bash
set -e
set -x

################################################################################
# Check if a PDF was scanned or created digitally, works on OCRd PDFs
#
# Usage:
#   bash is_scanned_pdf.sh [-p] file
#
#   Exit 0: Yes, file is a scanned PDF
#   Exit 99: No, file was created digitally
#
# Arguments:
#   -p or --pages: pos. integer, only consider first N pages
#
# Please report issues at https://github.com/jfilter/pdf-scripts/issues
#
# GPLv3, Copyright (c) 2020 Johannes Filter
################################################################################

# parse arguments
# h/t https://stackoverflow.com/a/33826763/4028896
max_pages=-1
# skip over positional argument of the file(s), thus -gt 1
while [[ "$#" -gt 1 ]]; do
  case $1 in
  -p | --pages)
    max_pages="$2"
    shift
    ;;
  *)
    echo "Unknown parameter passed: $1"
    exit 1
    ;;
  esac
  shift
done

# increment to make it easier with page numbering
max_pages=$((max_pages++))

command_exists() {
  if ! [ -x $($(command -v $1 &>/dev/null)) ]; then
    echo $(error: $1 is not installed.) >&2
    exit 1
  fi
}

command_exists mutool && command_exists gs && command_exists compare
command_exists pdfinfo

orig=$PWD
num_pages=$(pdfinfo $1 | grep Pages | awk '{print $2}')

echo $num_pages

echo $max_pages

if ((($max_pages > 1) && ($max_pages < $num_pages))); then
  num_pages=$max_pages
fi

cd $(mktemp -d)

for ((i = 1; i <= num_pages; i++)); do
  mkdir -p output/$i && echo $i
done

# important to filter text on output of GS (tmp1), cuz GS alters input PDF...
gs -o tmp1.pdf -sDEVICE=pdfwrite -dLastPage=$num_pages $1 &>/dev/null
gs -o tmp2.pdf -sDEVICE=pdfwrite -dFILTERTEXT tmp1.pdf &>/dev/null
mutool convert -o output/%d/1.png tmp1.pdf 2>/dev/null
mutool convert -o output/%d/2.png tmp2.pdf 2>/dev/null

for ((i = 1; i <= num_pages; i++)); do
  echo $i
  # difference in pixels, if 0 there are the same pictures
  # discard diff image
  if ! compare -metric AE output/$i/1.png output/$i/2.png null: 2>&1; then
    echo " pixels difference, not a scanned PDF, mismatch on page $i"
    exit 99
  fi
done

Solution 5:

There is no fail-proof way as far as I know, but there are some strategies.

The pdf may have some text embedded, but that may not be the text you are looking for. For example, some publishing companies, such as Jstor, put some text information on pdfs related to copyright even when the pdf is not ocred.

So a good strategy is to feed the pdf to a pdf to txt converter and count the number of words. If the number is too low (totally subjective metric) then it can be reasonably expected that it does not have ocr.

Below we have a bash one liner that mimics this for several files (parallel and poppler required):

find path/to/files/ -name "*.pdf" | parallel --progress -P3 pdftotext {} - | wc -w >> file.txt

This will count the number of words for each pdf file you have in your directory. You can then filter the ones that are below, e.g., 100 words, and feed then to a script such as ocrmypdf to ocr them.