To turn one big scanned PDF into separate, correctly named files, first split it where each document ends, using separator sheets with a barcode, blank pages, or a fixed page count. Then read a key piece of text from each document, such as a client name, date or invoice number, and build the file name and folder from it. Finally, review every proposed name before anything is saved, because OCR and barcodes can misread. You can do parts of this by hand in a PDF editor, with free tools, with an automation flow, or with a desktop app that does all three steps on your PC.
This guide explains why scans arrive as one file, how to choose a split method, how to name files from page text, which scan settings make reading reliable, how to file into folders automatically, how to make scans searchable offline, and how the common tools compare. Each section links to a deeper guide on this site.
Why does a stack of paper become one giant PDF?
Office copiers and document scanners are built for speed. You put a stack in the feeder, press one button and get one PDF, usually with a name like SCAN_20261028_0915.pdf, saved to a shared folder or emailed to you. That is fine for a single document. For a stack of 40 invoices, statements or letters, it means one file that someone has to split, rename and file by hand.
You can scan each document separately, but that moves the work to the copier, one button press at a time. The faster approach is to scan the whole stack at once and split it on the PC afterwards. Copier scan to folder: split one scan into named PDFs walks through setting up that flow.
Separator sheets, blank pages or every N pages: which split fits?
There are three common ways to tell software where one document ends and the next begins.
| Method | How it works | Best for | Watch for |
|---|---|---|---|
| Barcode separator sheets | A printed sheet with a barcode (such as Code 128 or QR) goes in front of each document | Mixed stacks with documents of different lengths | Printing and inserting sheets; keep them clean and flat |
| Blank pages | A blank sheet goes between documents | Simple stacks, no printing needed | Blank backs in duplex scans and near-blank pages can cause false splits |
| Every N pages | Split after a fixed page count | Stacks where every document has the same length, such as one-page forms | One missing or extra page shifts everything after it |
Separator sheets are the most reliable for mixed paperwork because the barcode is unambiguous. Some production scanners use patch codes instead, a different kind of separator read by the scanner itself; patch codes vs barcode separator sheets explains the difference. The guides on splitting by barcode separator sheets and splitting a PDF at blank pages cover each method in depth, and split a scanned PDF into separate files on Windows compares all three.
How do I name files from text on the page?
Most documents carry the information you would put in a file name somewhere predictable: a client name in the address block, an invoice number in the top right, a statement date under the logo. Zone OCR uses that.
- Draw a zone around the area that holds the reference text on a sample page.
- Read the zone with OCR on every document in the batch.
- Pull out fields from the text with a pattern, such as a date, a client name, a document type and a number.
- Check each field. A date should be a real date; an account number should have the right length.
- Build the name from a template, such as
{Client}\{Date}_{Type}_{Id}.pdf.
A consistent naming convention makes files sort and search well. Dates in year-month-day form, such as 2026-10-28, sort in date order. Avoid characters Windows does not allow in file names (\ / : * ? " < > |). File naming convention for scanned documents has examples for common document types, and how to rename scanned PDF files from an OCR zone is the step-by-step version.
Which scan settings make OCR and barcodes reliable?
Most copier defaults are tuned for small files, not for reading. A few changes at the panel, made once, prevent most misreads.
- Resolution: 300 dpi for normal office text; 400 dpi or more for small print such as receipts.
- Colour mode: grayscale rather than pure black and white when originals are faint, coloured or on textured paper.
- Size: scan at actual size, without “fit to page” scaling.
- Simplex or duplex: choose deliberately. Duplex scans of single-sided pages produce blank backs, which matter if you split at blank pages.
- Separator sheets: print barcodes at a decent size and keep the sheets uncreased.
Best scan settings for OCR and barcode splitting cites vendor guidance on resolution and explains the trade-offs with file size.
How do I file documents into client and date folders automatically?
Once each document has fields, the folder can come from the same template as the name. A template such as {Client}\{Year}\{Date}_{Type}.pdf files each document under its client and year, creating folders as needed.
Common setups:
- Bookkeeping: client folder, then year, with statements and receipts named by date and type. See split scanned invoices into separate PDFs for bookkeeping and split scanned bank statements, named by account and month.
- Property management: property, then vendor or tenant.
- Law offices: client and matter, then correspondence by date.
Before you shred the paper, check the record-keeping rules that apply to you. Tax agencies have conditions for relying on scanned records; shred receipts after scanning? CRA and IRS rules summarizes them. That guide is general information, not tax advice.
How do I make scanned PDFs searchable offline?
A scanned PDF is a picture of a page. To search it, it needs a hidden text layer added by OCR. Many tools do this by uploading pages to a cloud service. To keep files on your PC, use local OCR:
- Windows OCR. Windows includes a text recognition engine. It needs the OCR feature for each language you read; you can add a language in Settings > Time & language > Language & region. Apps that use Windows OCR then work offline.
- Free tools. NAPS2 is a free, open-source scanning app that can produce searchable PDFs with OCR. OCRmyPDF is a free, open-source command-line tool for adding a text layer to existing PDFs.
- PDF editors such as Acrobat Pro include OCR for scanned pages.
A text layer also helps beyond search: you can copy text out of the PDF, and screen readers can read it. Check that the tool keeps the page image unchanged and only adds invisible text, so the file still looks exactly like the paper.
Whichever you use, OCR text is only as good as the scan, so the settings in the section above matter for searching too.
How do I check names before anything is written?
A small example shows why. Suppose a stack of 30 vendor invoices is split and named from the vendor and invoice number. Twenty-eight names read cleanly. One invoice has a coffee stain over the number, and one vendor prints its name in a script font. Without review, those two land in folders nobody will look in. With review, they wait in a queue with the zone image beside them, and fixing them takes a minute.
Automated naming saves time only if mistakes are caught before files are saved. OCR confuses similar characters, such as 0 and O or 1 and l, and a misread client name files a document in the wrong folder, where it may stay lost.
Good practice:
- Preview every proposed name and folder before any file is written.
- Hold uncertain reads for review: reads that fail their checks, missing separators, duplicate names and unreadable files should wait for a person, with the zone crop shown so you can type the right value.
- Never touch the originals. Keep the combined scan until the split files are checked.
- Keep a record of what was written, ideally with a hash of each file, so you can show later that a file has not changed.
Tools compared: Acrobat, NAPS2, PaperPort, Power Automate and desktop apps
| Tool | Cost model | Splits scans | Names from page text | Notes |
|---|---|---|---|---|
| Adobe Acrobat Pro | Subscription | By number of pages, file size or top-level bookmarks | Naming convention for the split files, not read from the page | Full PDF editor with OCR; fine for fixed-length forms |
| NAPS2 | Free, open source | Starts a new file at Patch-T separator pages while scanning | Not its focus | Drives the scanner directly; strong free OCR |
| Tungsten PaperPort | Paid editions | Professional edition can pull apart, reorder and merge pages | Not its focus | A desktop document manager, a broader tool |
| Power Automate Desktop | Free for attended desktop flows on your PC; paid plans add premium features | With a flow you build, using its PDF page actions | With OCR actions you configure | Flexible; you design the page logic and error handling |
| ScanRoute Local | One-time purchase, with a free trial | Barcode, blank page or every N pages | Yes, one zone per document, with review | Splitting, naming and filing only |
Fair head-to-head guides: Acrobat Split PDF vs ScanRoute for scanned batches, NAPS2 vs ScanRoute, PaperPort vs ScanRoute and Power Automate Desktop vs ScanRoute.
How ScanRoute Local fixes this
ScanRoute Local turns one combined scan into separate, named PDFs, filed in the folders you choose, on your PC.
- Split three ways. Drop combined PDFs on the window, open them or watch a folder. It splits at Code 128 or QR separator sheets, at blank pages, or every N pages.
- Name from the page. Draw one zone over the reference line. It reads the text with the Windows OCR language installed on your PC, a pattern pulls fields such as type, date, client and number and checks each one, and a template such as
{Client}\{Date}_{Type}_{Id}.pdfbuilds each name and folder. - Nothing written until you commit. Preview lists every proposed name and folder. Documents that need a decision wait in an inbox with their zone crop: failed reads, a missing separator, a bad file name, a duplicate, an unreadable PDF or no OCR language.
- Originals only read. They are never changed, moved or deleted, and no file is overwritten. Each commit writes a CSV manifest listing every file written with its SHA-256 hash.
- Searchable output. When an OCR language is installed, upright scanned pages without text get an invisible text layer.
- Local. No account, no telemetry and no network connection.
Good to know: it works with PDF files in folders on your PC and reads one zone per document. It is not a scanner app or a document-management server. Without an OCR language, splitting still works and the fields wait for you to type them.
ScanRoute Local has a free trial on the Microsoft Store, then a one-time purchase.
Get ScanRoute Local from the Microsoft Store
Frequently asked questions
What is the most reliable way to split a mixed stack of scans?
Barcode separator sheets in front of each document. Blank pages work for simple stacks but can misfire on duplex scans.
Can OCR name my scanned files automatically?
Yes, by reading a zone on each page and building the name from a template. Review the names before saving, because OCR can misread.
What resolution should I scan at for OCR?
300 dpi for normal office text, and 400 dpi or more for small print. Grayscale helps with faint or coloured originals.
Can I make scanned PDFs searchable without uploading them?
Yes. Windows OCR, NAPS2 and OCRmyPDF can all add a text layer on your own PC.
More guides in this topic
- Split a scanned PDF into separate files on Windows
- Split a scanned PDF by barcode separator sheets
- Split a PDF at blank pages: how blank-page separators work
- Patch codes vs barcode separator sheets for scanning
- How to rename scanned PDF files from an OCR zone
- File naming convention for scanned documents
- Best scan settings for OCR and barcode splitting
- Copier scan to folder: split one scan into named PDFs
- Split scanned invoices into separate PDFs for bookkeeping
- Split scanned bank statements, named by account and month
- Shred receipts after scanning? CRA and IRS rules
- Acrobat Split PDF vs ScanRoute for scanned batches
- NAPS2 vs ScanRoute: split scans into named PDFs
- PaperPort vs ScanRoute: split and name scanned PDFs
- Power Automate Desktop vs ScanRoute for scanned PDFs
- All ScanRoute Local topics