Learn / OCR

OCR PDF Accuracy Tips

Getting garbled text from OCR? These practical tips will help you get near-perfect results from even tricky scanned documents.

Ready to run OCR? Use the free scanner tool.

Run OCR

What Affects OCR Accuracy?

OCR accuracy is determined mostly by the quality of the image going in. Even the best OCR engine can't reliably read blurry, skewed, or low-contrast text. Here's what matters most:

  • Resolution (DPI) — the single biggest factor
  • Page alignment — straight vs. tilted/crooked
  • Contrast — dark text on light background vs. faded or shadowed
  • Font type — standard printed fonts vs. handwriting or decorative fonts
  • Language setting — matching the document's language

7 Tips to Improve OCR Accuracy

1

Scan at 300 DPI minimum

Most scanner apps default to 150 or 200 DPI — change this before scanning. For documents with small fonts (legal footnotes, fine print), use 400 DPI. Going above 600 DPI adds file size without improving accuracy.

2

Keep pages flat and straight

A 5-degree tilt reduces OCR accuracy noticeably. Use a flatbed scanner rather than a phone camera when possible. If you're working with an existing crooked scan, use the Rotate PDF tool to correct the angle before running OCR.

3

Ensure good contrast

Dark text on a white background is ideal. Faded documents, colored paper, or light pencil marks reduce accuracy. If you're re-scanning, increase the scanner's contrast setting.

4

Clean up phone photos first

Phone cameras introduce shadows, perspective distortion, and glare that hurt OCR. Run your photo through PDF.it's Phone Scan Cleanup before OCR — it removes shadows, corrects perspective, and boosts contrast automatically.

5

Select the correct language

OCR engines use language models to resolve ambiguous characters. Selecting the right language (especially for non-English documents) can improve accuracy by 5–15%. This matters most for accented characters in Spanish, French, Portuguese, and German.

6

Split large PDFs for faster processing

Very large scanned PDFs (100+ pages) can time out. Split the PDF into smaller sections first using the Split PDF tool, OCR each section, then merge the results back.

7

Try grayscale instead of color for text documents

For black-and-white text documents, grayscale scans are smaller and process faster without sacrificing accuracy. Only use color scanning if the document has colored text or important colored graphics.

OCR Accuracy by Document Type

Document TypeTypical AccuracyBest Practice
Printed text, 300+ DPI98–99%Ready to use
Phone photo, good lighting90–95%Run Phone Scan Cleanup first
Old/faded typewritten document80–90%Boost contrast before scanning
Neat printed handwriting70–85%Manual review recommended
Cursive handwriting40–60%Manual transcription may be needed

Run OCR on Your Scanned PDF

Free OCR scanner — 10 conversions/day, no software required.

Run OCR Now

Frequently Asked Questions

What DPI should I scan at for best OCR accuracy?

300 DPI is the minimum recommended for OCR. 400–600 DPI is better for documents with small fonts. Above 600 DPI adds file size without meaningfully improving accuracy.

Why does OCR misread characters like 0 and O?

These characters look visually similar, especially at low resolution. OCR engines use language context to guess the right character — selecting the correct language in the OCR settings helps significantly.

Does color scanning improve OCR accuracy?

For standard black-and-white text, grayscale scanning is usually sufficient. Color scanning is helpful for colored text or colored backgrounds that would disappear in grayscale.

How do I improve OCR on a phone photo of a document?

Use PDF.it's Phone Scan Cleanup tool before running OCR. It removes shadows, flattens the perspective, and improves contrast — all of which significantly improve OCR accuracy.

Can OCR recognize handwriting?

Standard OCR is optimized for printed text. It works for neat, block-style writing but struggles with cursive or informal handwriting.