Convert Scanned PDF Tables to Excel with OCR

Turn a readable scanned PDF into an editable spreadsheet without retyping every cell. Upload one PDF, let DocToTable detect its table structure with OCR, review the preview, then export CSV or XLSX.

Upload a scanned PDF to DocToTable

Convert a scanned table now

1. Upload one PDF. Files can be up to 10 MiB and 500 pages; one conversion request processes up to 30 pages.
2. Wait while OCR reads scanned pages and the table detector identifies rows and columns.
3. Review the preview, paying particular attention to totals, dates, decimal separators, and repeated headers.
4. Download the result as Excel or CSV and make any document-specific corrections in your spreadsheet editor.

Without an account, a browser gets three finite previews that process the first two pages each; they do not reset daily. Eligible accounts may receive 10 one-time page credits, and each newly processed page uses one credit.

Scan checklist before upload

• Prefer a straight, evenly lit scan at about 300 DPI.
• Keep table text sharp and gridlines visible where possible.
• Remove shadows, page curl, and backgrounds that overlap the table.
• Use the original digital PDF when one exists; OCR is for image-based pages.
• Split unrelated documents before upload so one PDF contains the table you want.

For detailed explanations of resolution, contrast, skew, and OCR troubleshooting, read the OCR table extraction guide.

What to verify in the preview

Check that the first row is the intended header, columns contain the same type of data, and page headers or footers have not become data rows. The web product detects columns automatically; adjustments such as renaming columns or removing repeated rows happen after export.

For the broader native-and-scanned PDF workflow, see how to convert PDF tables to Excel.

Common scanned-table jobs

• Invoice line items: verify quantities, unit prices, tax, and totals.
• Statements: compare opening and closing balances and check date formats.
• Research appendices: validate special symbols and repeated headers.
• Archived records: spot-check names, identifiers, and faint characters.

OCR quality depends on the source. A clean preview is a review aid, not a guarantee; reconcile critical values against the PDF before using the spreadsheet downstream.