Urdu poetry OCR: ghazals and nazms with their line breaks
Digitise diwans, ghazals and nazms with each misra on its own line, in Nastaliq, ready to edit, publish or translate.
The problem
- Poetry books mix centred couplets, headings and footnotes on one page.
- A single dot changes a word, and a changed word breaks the metre.
How LexaOCR handles it
- Verse is recognised and kept line by line, never run together into prose.
- Every word is voted on by several readers and checked against an Urdu dictionary; doubtful words are shown to you.
- Send the finished text to LexaTranslate for a cultural translation of the poetry.
What you get
- A Word file with couplets laid out as verse
- A review edition for proofreading
- An audio reading of the poems
Questions
Does it understand the metre?
No. LexaOCR reads the words faithfully; it does not judge metre. That is why every uncertain word is shown to you.
Try it on one page
The quality check is free, and nothing is read until you confirm.