Quote:
Originally Posted by willus
So I had guessed based on what markom wrote in an earlier post that duokan uses OCR to re-flow, but now it seems like maybe it doesn't? Does it do graphical text re-flow on scanned text a la k2pdfopt (no OCR), or is it doing real-time OCR? How long has it been able to re-flow scanned text?
|
It uses graphical reflow a la k2pdfopt only on pdf scans without ocr layer in the background.
It is very fast on iPad 4 and it almost instantly changes size of letters (about 8 possible levels) and distance between sentences(three levels).
For pdf scans with ocr layer in the background it is similar as for textual pdfs, Duokan reflows and shows ocr layer (pictures lost).