Urdu book OCR: scanned books to editable Word
Turn a scanned Urdu book, in Nastaliq or Naskh, into clean, editable text and a Word file laid out like the original, without retyping it.
The problem
- Most Urdu books were printed in Nastaliq, a sloping, connected script that general OCR tools read poorly.
- Retyping a 300-page book by hand takes weeks and introduces its own errors.
- Old paper, faded ink and tight line spacing make the dots and small letters easy to misread.
How LexaOCR handles it
- A free quality check scores every page and forecasts its accuracy before anything is read, so you can rescan weak pages first.
- Several independent readers, including one made for Nastaliq, read each line. Where they agree, the word is accepted.
- Disagreements are checked against a 146,000-word Urdu dictionary and, when needed, a linguist AI.
- You review only the words still in doubt, with each reader's version one keystroke away.
- Books up to 1 GB upload in pieces and keep processing on the server if you close your laptop.
What you get
- A clean Word edition of the whole book
- A review edition with each scanned page beside its text
- Searchable PDF and plain text
- Optional audiobook in a calm news-anchor voice
- One-click hand-off to LexaTranslate
Questions
Can LexaOCR read Noori Nastaliq books?
Yes. LexaOCR has a reader made for Urdu Nastaliq and checks every word against an Urdu dictionary. Accuracy depends on the print and scan quality, which the free quality check forecasts per page.
How long does a whole book take?
Pages are read in parallel on the server and saved one by one. You can close the page while it works, and LexaOCR can email you the Word files when it finishes.
Will the poetry keep its line breaks?
Yes. Verse is recognised and kept as separate lines, and headings and paragraphs keep their structure.
Try it on one page
The quality check is free, and nothing is read until you confirm.