Hello, Am Donnerstag, den 13.05.2010, 12:34 -0700 schrieb Tom: > That depends on many factors. Are you talking about dewarping of > camera-captured images, or dewarping of flatbed scanned pages? And > what purpose do you want to dewarp the images for?
Both, or in detail we are using camera based flatbed scanners where the book is pressed against a glass plate and scanned by a camera on top. Also we used classical flatbed scanners. > Generally, no simple global model is going to work well for either > situation. > We know and that is the reason to ask you :) > If this is just for OCR, we're currently integrating layout analysis > into OCRopus that can deal with some degree of curvature. With that, > you should be able to perform OCR on lightly distorted pages directly > without dewarping. We have developed a segmenter which could handle this. The problem is on OCR side. We are planning to use a lot of different engines and we think we should not trust them to do the job well. ;-) Bye Andreas -- Andreas Romeyke - Abteilung Blindenschrift - Deutsche Zentralbücherei für Blinde zu Leipzig (DZB) Gustav-Adolf-Straße 7, 04105 Leipzig Tel: +49 341 7113-..., Fax: +49 341 7113-125 Internet: www.dzb.de E-Mail: [email protected]
signature.asc
Description: Dies ist ein digital signierter Nachrichtenteil
