Summary: Internal Report 2/97, Albert­Ludwigs­Universit¨ at Freiburg, IIF­LMB, Germany, November 1997
Stereo Image Coding
S. Siggelkow
Albert­Ludwigs­Universit¨ at Freiburg
Institut f¨ ur Informatik
79085 Freiburg i. Br., Germany
sven.siggelkow@informatik.uni­freiburg.de \Lambda
Abstract. This paper discusses a method for stereo image coding. We assume
stereo images that try to imitate the human visual system, i.e. the two camera
planes must be coplanar. First we apply an intensity based stereo algorithm us­
ing cooperative bidirectional matching with a hierarchical multilevel structure that
returns depth field and occlusion information. This information can be used to esti­
mate the disparity of pixels from one image to the other one. On the one hand one
might use the full depth information to be able to generate synthetic virtual views,
on the other hand, if one is only interested in transferring the stereo image pair, one
can divide the image into objects of similar depth and code these like in MPEG­4.
Similar to objects in image sequences, the object of the second stereo image can be
predicted by applying a translation to each object.
1 Introduction
For future telecommunication technologies a highly effective interpersonal information ex­

Source: Albert-Ludwigs-Universität Freiburg, Institut für Informatik,, Lehrstuhls für Mustererkennung und Bildverarbeitung

Collections: Computer Technologies and Information Sciences