Photo or screenshot
Upload a PNG, JPG, or WebP image with a visible table.
Extract rows and columns from a photo or screenshot of a table, correct the cells, and download the result as XLSX or CSV.
PNG, JPG, or WebP, up to 20 MB. For best results, use a sharp image taken straight on.
Florence-2 can return OCR text along with the coordinates of each region. UtiliTools Pro takes advantage of these positions to estimate rows and columns, and then presents the result as a table that you can correct before exporting it.
Upload a PNG, JPG, or WebP image with a visible table.
Read text and its position with Florence-2.
Automatically organize content into rows and columns.
Correct the content before exporting.
Generate an Excel-compatible spreadsheet.
Download the data in a simple and universal format.
Select a photo or screenshot containing a table.
Click Extract table with AI to recognize text and its positions.
Check the rows and columns and correct the necessary cells.
Download the result as Excel XLSX or CSV.
Unlike plain-text OCR, this tool uses Florence-2โs OCR with regions task. Each recognized text fragment includes coordinates, which help reconstruct the layout of a table in the image.
Merged cells, tilted photos, faint borders, or tightly spaced text can lead to incorrect columns. You can edit the resulting table directly before downloading.
Yes. The tool uses regional OCR to recognize text and its positions, reconstruct an editable table, and let you download it as XLSX.
Yes. After checking the cells you can download both Excel XLSX and CSV.
No. Accuracy depends on image quality and table structure. Sharp, straight-on images with well-aligned columns usually work best. Review the result before downloading.
Yes. You can edit values and add or remove rows and columns in the preview.
No. The OCR runs locally in the browser after downloading the files of the Florence-2 model.