We are aiming at extending the basic digital camera functionalities to the ability to simulate the flattening of a document, by virtually acting like a flatbed scanner. Typically, the document is the warped page of an opened book. The problem is stated as a computer vision problem, whose resolution involves, in particular, a 3D reconstruction technique, namely shape from contour. Assuming that a photograph is taken by a camera in arbitrary position or orientation, and that the model of the document surface is a generalized cylinder, we show how the corrections of its geometric distortions, including perspective distortion, can be achieved from a single view of the document. The performances of the proposed technique are assessed and illustrated through experiments on real images.
Shape from Contour for the Digitization of Curved Documents
Asian Conference on Computer Vision ; 2007 ; Tokyo, Japan November 18, 2007 - November 22, 2007
01.01.2007
10 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
IEEE Computer Society , Frontal View , Image Point , Document Image , IEEE Conf Computer Science , Image Processing and Computer Vision , Computer Imaging, Vision, Pattern Recognition and Graphics , Pattern Recognition , Artificial Intelligence , Biometrics , Algorithm Analysis and Problem Complexity
Shape from shading for the digitization of curved documents
British Library Online Contents | 2007
|Reconstruction of torn documents using contour maps
IEEE | 2005
|Flattening curved documents in images
IEEE | 2005
|Reconstruction of Torn Documents using Contour Maps
British Library Conference Proceedings | 2005
|Springer Verlag | 2018
|