Deskew PDF
Detect and correct the tilt in scanned PDF pages automatically. Straighter pages look better and give measurably better OCR results.
Upload Scanned PDF Files
About This Tool
Paper never goes through a scanner perfectly square. A degree or two of tilt is normal, invisible on any single page, and obvious across forty of them - the document reads as sloppy even though nothing is wrong with the content.
It also costs you accuracy. OCR engines expect horizontal text lines, and recognition rates fall measurably on skewed pages. Deskewing before OCR is one of the cheapest accuracy improvements available.
Detection is automatic and per page, because different sheets in the same batch tilt by different amounts. The angle is measured from the dominant text baselines, and each page is corrected by its own value. Up to 10 files can be processed in one run.
How to Use
Add your scanned PDFs
Drop in up to 10 files.
Let it detect the angles
Each page is measured independently from its text baselines.
Review the corrections
Detected angles are listed per page so you can spot any that look wrong.
Apply and download
Save the straightened document.
Use Cases
Cleaning up a batch scan
Forty pages, each tilted differently, corrected in one pass.
Preparing for OCR
Straightening first measurably improves recognition accuracy.
Archival scanning
Digitised records that look deliberate rather than rushed.
Frequently Asked Questions
How is the angle detected?
From the dominant text baselines on each page, measured page by page since tilt varies across a batch.
What if it gets a page wrong?
Pages dominated by images or unusual layouts can mislead the detector. Use Rotate Custom to set the angle by hand for those.
Does correcting the tilt degrade the image?
Rotation resamples raster content, so there is a very slight softening. It is far outweighed by the readability and OCR gains.
Should I deskew before or after OCR?
Before, always. OCR accuracy depends on horizontal text lines.