Author/Editor		Ude, A; Ekre, TE
Title		Stereo grouping for model-based recognition
Type		članek
Source		In: Proceedings of the 13th international conference of pattern recognition. Track 1: computer vision; 1996; Vienna,
Publication year		1996
Volume		str. 1-2
Language		eng
Abstract		A strategy for the fusion of information from a stereo image pair for model-based object recognition is discussed. Our scheme combines a new method for feature grouping with a region-based stereo matching and a hypothesize-and-verify paradigm. The developed grouping method is based on a graph theoretical algorithm. It exploits prior knowledge to find the groups of image features which are likely to come from a sought model(s). Bayesian classification is used to order the resulting hypotheses. A mechanism for a dynamic threshold modification is incorporated into the system to enable the grouping at different resolutions. Unlike classical techniques for object recognition from stereo, our strategy does not depend on a data driven computation of a depth map. We argue that a purposive reconstruction of 3-D information can be more efficient and robust.
Descriptors		DATABASES, BIBLIOGRAPHIC AUTOMATIC DATA PROCESSING