PDF & Image Table Extractor

1. Drop a PDF or image file into the tool, or click browse file. The picker accepts one file at a time.
2. For a PDF, choose a page with PDF page. Draw a rectangle around the table in the preview.
3. Use Whole page as region when the table fills the page. Try Auto-detect region when the table has visible ruling lines.
4. Click Extract, then inspect the reconstructed table and edit any cells that need correction.

How PDF table extraction reads text

For a PDF, the tool checks the selected region for text items. With at least two, it builds the table from the PDF text layer. Image files and PDF regions with fewer use OCR. Checking Force OCR sends the selected region through OCR even when PDF text is available. OCR uses the English eng model.

Ruling lines define row and column boundaries when enough horizontal and vertical lines are found. Without them, the tool groups text by row position and whitespace gaps. An unusual layout can place text in a nearby inferred row or column. The preview gives each cell a confidence percentage based on its assigned words.

For example, a table can become this CSV:

Product | Qty
Pens | 12
Folders | 4

Product,Qty
Pens,12
Folders,4

How to edit and export extracted tables

Edit cell text directly in the preview. First row is header starts enabled, so turn it off when the first row is ordinary data. To create a merged cell, Shift-click two or more cells and choose Merge selected. Alt-click a merged cell to split it.

• CSV and Markdown flatten a merged value by repeating it across the cells it covers.
• XLSX keeps manually created merge spans as spreadsheet merges.
• JSON includes cell values, confidence, table structure, and region coordinates.

Each table has buttons for CSV, XLSX, JSON, and Markdown. The combined controls export all tables as a multi-sheet XLSX workbook, one JSON file, or a ZIP containing CSV and Markdown files. Use Delete table to remove a result.

When a PDF table needs a tighter region

Auto-detect region looks for ruling lines. When it finds them, it selects their bounding area. Otherwise, it selects the whole page or image as a starting region. Draw a smaller rectangle around the table before clicking Extract when the starting region is too large. Use Clear regions on this page to start over, and Change file to load another input.

The tool exposes the first 500 pages of a PDF. Password-protected or encrypted PDFs are rejected. The file is read and processed in your browser, and its contents are not sent to a server.

Where this comes up

• Select a table in a scanned invoice, review cells with lower confidence, and export XLSX for a workbook.
• Select a table in a PDF appendix, export JSON, and keep the region coordinates with the extracted data.
• Select a schedule in a screenshot, correct an OCR heading, and export Markdown for a project note.

When to use PDF table extraction

Use PDF Extract Tools when you need to choose among PDF extraction tasks. Use Extract Text From PDF when you need page-by-page text, page ranges, OCR pages, or a searchable PDF instead of a reconstructed table.

Frequently asked questions