Document perspective correction
Turn an angled document photo into a rectangular view by marking its four corners.
How to correct document perspective in 3 steps
Open the photographed page
Choose a clear photo where the document corners are visible. Perspective correction starts from the image at this editing step.
Position the four corners
Use the drawing controls or corner X/Y percentages. Place corners clockwise from top-left around a convex quadrilateral, then choose automatic or explicit output dimensions.
Apply and inspect
Apply the correction and check page edges and text. Export the image, or open Images to PDF to assemble several corrected pages.
Straighten a flat page photographed at an angle
Four-corner correction maps a selected quadrilateral to a rectangle. Careful corner placement matters: crossing the corners creates an invalid shape, while inaccurate edges can stretch text. The tool works on perspective in a flat surface; a curved book page or lens distortion needs a different kind of correction.
Try it with your own files
Position four corners clockwise from top-left around a photographed document, then apply the perspective correction.
Know the limits
Corner placement is manual. There is no automatic detection or OCR. Strong lens distortion and curved pages are not corrected.
Private image editing on your device
Your images are opened and processed in your browser, not uploaded to a server. The website downloads the code and assets it needs to run, and some tools download a processing model. Those requests do not contain the photos you open. Download your results before closing the page; working images are not kept between visits.
Read our Privacy PolicyFrequently asked questions
Does it detect document corners automatically?
No. Corner placement is manual, using drawing controls or numeric coordinates.
Can it make blurry text readable?
It can straighten the page but cannot recover detail missing from a blurry or low-resolution source.
Will the result be an editable document?
No. The result remains an image. You can add it to a PDF, but this tool does not perform OCR.