Summary: Creating Generative Models from Range Images
Ravi Ramamoorthi
Stanford University #
ravir@graphics.stanford.edu
James Arvo
California Institute of Technology
arvo@cs.caltech.edu
Abstract
We describe a new approach for creating concise high­level gener­
ative models from range images or other approximate representa­
tions of real objects. Using data from a variety of acquisition tech­
niques and a user­defined class of models, our method produces a
compact object representation that is intuitive and easy to edit. The
algorithm has two inter­related phases: recognition, which chooses
an appropriate model within a user­specified hierarchy, and param­
eter estimation, which adjusts the model to best fit the data. Since
the approach is model­based, it is relatively insensitive to noise and
missing data. We describe practical heuristics for automatically
making tradeoffs between simplicity and accuracy to select the best
model in a given hierarchy. We also describe a general and efficient

Source: Arvo, Jim - Departments of Information and Computer Science & Electrical and Computer Engineering, University of California, Irvine
O'Brien, James F. - Department of Electrical Engineering and Computer Sciences, University of California at Berkeley
Stanford University - Computer Graphics Laboratory

Collections: Computer Technologies and Information Sciences