OCR
Old paper, new search
Docu Asia recognises text in scanned PDFs and images on your own computer, then indexes that text like any other document. Nothing is uploaded to a recognition service.
When OCR runs
- · A PDF page has little or no selectable text
- · You add a JPG, JPEG, PNG, TIFF or BMP image
- · An email attachment contains a scan, when image OCR is enabled
- · Pages with a digital text layer are skipped to save time
What you see
- · An OCR indicator on every result that came from a scan
- · Recognised text in the preview, with your terms highlighted
- · Page numbers kept where the source allows
- · A low-confidence warning when recognition is unreliable
Local only
Recognition runs in a background worker on your PC.
Multilingual
International and Indian languages, chosen by you.
Non-blocking
The interface stays usable while pages are recognised.
Honest results
Confidence is shown, not hidden.
Languages
Choose the languages that appear in your documents. Fewer languages give faster, more reliable recognition per page. Availability depends on the recognition data installed with the app.
Limits
Where OCR struggles
Good conditions
Straight pages at 300 dpi or more, printed text, clear contrast and one or two languages per page.
Difficult conditions
Handwriting, carbon copies, faint stamps, skewed or folded pages, mixed scripts and photographs of screens. Expect missing or garbled words in these cases.
Please note
- · OCR accuracy depends on image quality, resolution, font and the condition of the document.
- · Recognised text is an approximation of the page; always open the original for anything important.
- · Images and recognised text are never sent to a remote recognition service by default.
Find the document, even when you forgot the filename.
Search PDFs, scans, Office files, images and email archives directly on your Windows PC. Your documents stay on your computer.