Back to blog

How to Remove Text and Date Stamps from Images

Select captions, timestamps, and unwanted lettering by hand, then rebuild the covered background with AI.

Sep 5, 2026MagicEraser TeamMagicEraser Team
How to Remove Text and Date Stamps from Images

Removing a caption or date stamp is easiest when the selection follows the lettering closely and preserves enough nearby texture for reconstruction. MagicEraser uses manual mask tools for this job, so you decide exactly which text should disappear.

Open the Text Remover and use the workflow below for text you own or have permission to edit.

Text Remover upload area before choosing an image
Step 1: Upload the clearest source image so small characters and background texture remain visible.

Use the original image when possible

Start with the clearest JPG, PNG, or WEBP available. Small characters, thin outlines, compression artifacts, and translucent shadows are easier to select in a high-quality source. Avoid taking a screenshot of an image when the original file is available.

Before drawing, identify what lies underneath the text. A plain sky or wall is usually straightforward. Repeating fabric, product edges, faces, and detailed illustrations require a tighter mask and closer review.

Match the selection tool to the lettering

Use Rectangle for a horizontal caption, label, subtitle, or date stamp. Drag close to the top, bottom, and ends of the text instead of selecting the entire surrounding panel.

Rectangle tool selecting a line of unwanted text
Step 2: Keep a rectangular selection close to a straight caption or date stamp.

Use Lasso for angled lettering, curved baselines, perspective text, or words arranged around a shape. Trace the group closely while leaving nearby subject detail outside the selection.

Use Brush for small characters, punctuation, thin strokes, and missed corners. If the mask touches an area that should remain, switch to Eraser and subtract it. Undo and Redo make it easy to compare corrections.

The editor does not guess which text you intend to remove. This manual approach is useful when an image contains both wanted and unwanted labels.

Brush and Eraser tools refining a manual text mask
Step 3: Add missed strokes with Brush and subtract nearby subject details with Eraser.

Cover complete characters without making the mask oversized

Include every visible stroke, outline, glow, or shadow belonging to the unwanted text. A partial letter can leave a recognizable fragment after processing. At the same time, avoid selecting a much larger rectangle than necessary because that removes useful original background.

When several text groups are far apart or sit on different textures, process them separately. A date in the corner and a caption across a face are two different reconstruction problems.

Tune edge expansion and inspect the background

Edge expansion extends the mask slightly beyond its drawn boundary. Start with the default value. Increase it if a thin outline remains around solid lettering, or reduce it near product edges, hair, and other fine detail.

Manual text mask shown with the Edge Expansion slider
Step 4: Adjust Edge Expansion so letter outlines disappear without covering nearby fine detail.

After processing, compare Before and After at high zoom. Check straight lines, repeated tiles, table edges, fabric patterns, and gradients. If only one small artifact remains, choose Continue editing and mask that area in a fresh session.

Text removal Before and After result with continue and download actions
Step 5: Compare the restored surface, then download or continue editing only the remaining artifact.

Signed-in users may also return to successful results from History. An account is optional for the initial upload, mask, erase, and download workflow.

Common reasons text removal needs another pass

  • The mask missed a letter outline, shadow, glow, or punctuation mark.
  • A large rectangle removed too much useful background around the text.
  • Several distant text groups were processed as one reconstruction area.
  • The source was heavily compressed or the lettering was too small to select accurately.
  • The text crossed a face, product boundary, or complex repeating pattern.

For an authorized text watermark, use the same careful selection principles in the Watermark Remover. Only modify images you own or have explicit permission to edit, and respect the rights of the original creator.