LocalKit
Visit LocalKit.app
Local-first OCR Tool

Table Extractor

Upload a screenshot or scan containing rows and columns to instantly detect cell breaks, preview the table, and export to CSV.

Capture tables from scans without retyping columns. Accepts images (JPG, PNG, HEIC) and PDFs.

Drag & drop supported. Processing happens locally in your browser; nothing is uploaded.

How to use Table Extractor

Whether you are digitizing a whiteboard grid or a scanned timesheet, this extractor rebuilds the table structure automatically. By boosting contrast and detecting whitespace gaps, it can infer columns even when dividers are faint, saving you from copying cells into spreadsheets by hand. The workflow honors existing alignment, keeps header casing intact, and surfaces any rows with suspicious gaps so you can fix them before exporting.

  1. Upload a screenshot, scan, or camera photo that contains a clear grid or rows and columns.
  2. Watch the OCR worker enhance the image and mark the lines. The status bar lets you know when parsing is complete.
  3. Review the interactive preview. Click any cell to adjust text, merge rows, or correct headings before export.
  4. Download the CSV to share with your team or continue editing in your spreadsheet tool of choice.

If your table has merged headers, crop the image so the relevant section fills the frame. The extractor reads edge-to-edge to preserve spacing, which makes the downstream CSV much easier to reconcile. For dense schedules, run a quick preview to verify columns before handing the file off to teammates who depend on trustworthy row counts.

FAQ

Do I need perfect grid lines?

Clear horizontal and vertical dividers improve the preview, but the extractor also uses whitespace detection for looser layouts.

Can I edit cells before export?

Yes. Adjust the preview inline and the CSV download will follow your edits.

Does it run offline?

The OCR runs in-browser. Once the page is loaded you can disconnect from the internet and continue extracting tables.

Related tools

PDF to TextInvoice to CSVSerial/Model Number Extractor