← All image tools

Perspective correction

Straighten photographed documents and correct trapezoid distortion.

Everything happens locally in your browser. Nothing is uploaded.
Perspective correction
Straighten photographed pages by compensating for trapezoid distortion.
Private browser processing
No uploadsNo accountFree export

A practical guide to Perspective correction

Perspective Correction reduces trapezoid distortion in a photographed page by progressively adjusting the left and right edges from top to bottom. The output stays local and is useful for making a slightly skewed document look straighter before sharing.

Example: straighten a photographed receipt

Choose the clearest receipt image, increase the top-edge correction until the page sides look more parallel, and inspect small print before using the exported result.

Task-specific steps

  1. 1

    Choose a document photo with visible page edges and enough margin that a correction will not cut into important content.

  2. 2

    Adjust the top-edge correction percentage conservatively while comparing the upper and lower edges of the page.

  3. 3

    Create the corrected image, inspect text and borders at full size, and retain the original for a stronger geometric correction later.

Options that affect the result

  • The inset slider controls how much narrower the top edge is treated relative to the lower edge.
  • The workflow preserves the image height while reducing the width to avoid inventing pixels beyond the source boundaries.

Limitations to know first

  • This lightweight correction is not a four-corner homography and cannot fully repair severe camera angle, rotation, or curved pages.
  • Cropping the side edges removes pixels, so extreme inset values can discard margins, seals, or text near the border.

Questions specific to this task

What kind of distortion does this tool correct?

It compensates for a simple trapezoid where the photographed top edge appears narrower than the bottom. Severe skew or four independently positioned corners need a document scanner or editor.

Why did the corrected image become narrower?

The workflow removes the side inset needed to compensate for the simulated perspective while keeping the original height and avoiding empty output regions.