Odia is served by far fewer tools than the larger Indian languages, which makes scanned Odia paperwork particularly awkward. ROR land records from Bhulekh, OSSC and BSE Odisha papers, TPCODL bills and tahasil notices are all images until an Odia model reads the rounded script.
When to use this
Common inputs: ROR records from Bhulekh Odisha, OSSC and OPSC papers, BSE Odisha mark sheets, TPCODL bills, tahasil and panchayat notices.
Frequently Asked Questions
Why do so few tools support Odia?
Odia has a smaller training corpus than Hindi or Tamil, so most commercial OCR skips it. The open Tesseract model used here does cover Odia, which is why this works at all, though accuracy on poor scans trails the larger languages.
Powered by OCR PDF.