Document Processing for Scans
Document Processing is a family of features—like OCR—that adds functionality and automation to scans or improve image quality:
Optical Character Recognition (OCR) is the process of taking an image, such as a scanned document, and reconstructing its text. This allows scanned documents to become searchable and/or editable.
Text-searchable documents have two major benefits over other scan outputs:
You can search for and copy specific content within the document.
If the document has been added to a document management system, you can find the document by searching for its content.
Performing OCR is a resource intensive process that can add seconds or tens of seconds per page to the time it takes to deliver a document. For this reason, enable OCR on scan actions where it is most useful, not where fast delivery is more important.
Currently PaperCut MF supports the following text-searchable file types:
PDF (text-searchable)—PDF v1.4 with PDF/A-1 compliance according to the requirements defined by the PDF/A standard.
OCR supports extracting text for approximately 100 languages. You can choose to use up to 10 of those languages, however for the best performance, limit your choices to a maximum of four languages.
A F L S Afrikaans Faroese Lao Sanskrit Albanian Filipino Latin Scottish Gaelic Amharic Finnish Latvian Serbian Arabic Flemish Letzeburgesch Sindhi Armenian Franksh Lithuanian Sinhala; Sinhalese Assamese French Luxembourgish Slovak Azerbaijani G M Slovenian B Gaelic Macedonian Spanish Basque Galician Malay Sundanese Belarusian Georgian Malayalam Swahili Bengali German Maltese Swedish Bosnian Greek Maldivian Syriac Breton Gujarati Maori T Bulgarian H Marathi Tagalog Burmese Haitian Moldavian Tajik C Haitian Creole Moldovan Tamil Catalan Hebrew Mongolian Tatar Cebuano Hindi N Telugu Cental Khmer Hungarian Nepali Thai Cherokee I Northern Kurdish Tibetan Chinese - Simplified Icelandic Norwegian Tigrinya Chinese - Traditional Indonesian Occitan (post 1500) Tonga (Tonga Islands) Corsican Inuktitut Oriya Turkish Croatian Irish P U Czech Italian Panjabi Uighur D J Pashto Persian Ukrainian Danish Japanese Pilipino Urdu Dhivehi Javanese Polish Uyghur Divehi K Portuguese Uzbek Dutch Kannada Punjabi V Dzongkha Kirghiz;Kyrgyz Pushto Valencian E Kazakh Q Vietnamese English Korean Quechua W Esperanto Kurdish R Welsh Estonian Romanian Western Frisian Russian Y Yiddish Yoruba
Batch Splitting transforms a single large input document into multiple output documents. It’s ideal for use with high-capacity document feeders, and when scanning batches of forms or invoices.
Batch Splitting is applied at the scan action level, and has two options:
- Split every N pages (set by the administrator)
- Split on blank separator pages
When splitting on blank separator pages, a detected blank page becomes the last page of the current document and the following non-blank page becomes the first page of the next document.
Output documents all share the same image and quality settings like DPI, color, and orientation. And they are all delivered to the same destination with the same root file name appended numerically for each document, like this:
You can use splitting on blank pages in combination with Blank Page Removal. The blank pages are first detected to trigger splitting, then removed from the resulting individual documents.
Blank Page Removal detects and removes pages with no content. This helps reduce the size of scanned documents and provides a more seamless experience when reading the digital output.
You configure Blank Page Removal ON or OFF at the scan action level. It defaults to OFF.
It works by analyzing each page against a white space threshold, and deletes pages which meet or exceed the threshold. This threshold is set by default, but you can manually adjust it using the
system.scan.docproc.blank-threshold configuration key to fine-tune sensitivity.
PaperCut MF applies Blank Page Removal at the page level, not sheet level, so it is ideal for shrinking scans of single-sided content.
When there are multiple blank pages in a row, they are treated as one blank page and are removed together.
These settings are applied globally, from the Options > Capture page. When activated, they apply to all new and existing scan actions with no additional setup required.
Despeckle detects pixel noise in the scanned image and removes it. This is useful when scanning documents that have already been copied or printed multiple times, like student forms, or when scanning documents using older imaging hardware.
Deskew detects crooked documents, and straightens them by up to 45 degrees. This type of straightening is most helpful when documents are being scanned individually on the glass, where there’s a high risk of human error or paper shifting alignment as the lid is opened and closed.
Both Deskew and Despeckle help improve the accuracy of OCR, and can be used individually or in combination with any other Document Processing features.
PaperCut MF provides the ability to run Document Processing using the PaperCut MF Cloud Document Processing service (formerly MF Cloud OCR Service), (one of PaperCut’s Cloud Services), or using your own self-hosted infrastructure:
PaperCut Cloud Document Processing service — Let the cloud do all the heavy lifting and benefit from:
improved local infrastructure performance
automatically deployed service updates. Always have the latest performance improvements and functionality.
The PaperCut MF Cloud Document Processing service processes concurrent jobs in parallel and handles any scaling of the service, even when there is a high user load.
Self-hosted Document Processing (On-prem) — For use when there’s a requirement to host Document Processing on your organization’s local infrastructure, and you have a high performing Application Server or resources to configure multiple Document Processing servers.
Some organizations have a requirement for data to stay within their own managed infrastructure or even on their own premises, typically for regulatory or compliance reasons. Be aware that this involves installing the service on selected infrastructure and keeping it updated by installing new versions. For more information, take a look at the Document Processing FAQs, or to get started, see Set up self-hosted Document Processing