How to convert a photo of a table to Excel
Last updated September 2026
To convert a photo of a table to Excel, photograph it straight-on in even light, then upload it to an AI table extraction tool that reads the grid and rebuilds it as rows and columns. Photos are harder to read accurately than screenshots because of glare, angle, and lighting, so review every cell against the original before exporting — especially numbers.
- 1
Shoot straight-on, not at an angle
Perspective skew distorts the grid the AI is trying to read. Hold the phone directly above or in front of the table, not tilted.
- 2
Use even light — avoid a hard shadow
A shadow cutting across half the table is one of the most common causes of missed cells. Natural, diffuse light beats a single overhead bulb.
- 3
Fill the frame with the table
Get close enough that the table takes up most of the shot, without cropping into its edges — losing an edge column is the single most common mistake.
- 4
Upload the photo to an extraction tool
The AI locates the grid even with some skew or uneven lighting, and reads it into rows and columns within seconds.
- 5
Check every cell against the original photo
Photos score lower confidence than screenshots on average — a faded price, a smudged digit, or a shadowed corner are the usual suspects.
- 6
Export a real .xlsx workbook
With numbers typed as actual numbers, not text, so totals and sorting work the moment the file opens.
A real example: a printed inventory sheet
A warehouse has a printed stock count sheet clipped to a board — no scanner nearby, no time to retype forty rows by hand. A phone photo, shot straight-on under the warehouse’s overhead lighting, is the fastest path from paper to a usable file.
The extraction handles the printed typeface reliably, but flags three cells in a column where the paper had creased and partly obscured a digit. Checking those three against the original photo takes under a minute, and the corrected table exports as .xlsx — ready to reconcile against the previous count.
Frequently asked
A screenshot is pixel-perfect digital text with no lighting or angle to distort it. A photo introduces glare, shadow, and perspective skew, all of which lower confidence on individual cells.
It attempts handwriting, but accuracy is meaningfully lower than printed text — plan to double-check every cell rather than a few flagged ones.
Moderate blur is often still readable, but a genuinely out-of-focus photo is worth retaking — accuracy tracks image clarity closely.
Yes — that’s the specific case this solves. A phone camera photo works the same way a scan would, just with more variables to control for.