Geometric Reasoning in a Parallel Network
- 1 April 1991
- journal article
- Published by SAGE Publications in The International Journal of Robotics Research
- Vol. 10 (2) , 103-122
- https://doi.org/10.1177/027836499101000203
Abstract
The general principles of geometric reasoning for computer vision are discussed and some previous attempts at imple menting reasoners reviewed. A new implementation is pro posed that involves interval arithmetic to cope with uncer tainties and has a network structure that promotes efficiency from parallelism and accuracy from feedback. Typical con straints from a 3D model/3D image vision system are dis cussed and used to evaluate the network performance.Keywords
This publication has 11 references indexed in Scilit:
- Optimal Combination and Constraints for Geometrical Sensor DataThe International Journal of Robotics Research, 1988
- Sensor Models and Multisensor IntegrationThe International Journal of Robotics Research, 1988
- Building, Registrating, and Fusing Noisy Visual MapsThe International Journal of Robotics Research, 1988
- Solving geometric constraints in a parallel networkImage and Vision Computing, 1988
- Uncertain geometry in roboticsIEEE Journal on Robotics and Automation, 1988
- Geometric reasoning for computer visionImage and Vision Computing, 1987
- Constraint propagation with interval labelsArtificial Intelligence, 1987
- SMS: a suggestive modelling system for object recognitionImage and Vision Computing, 1987
- Symbolic reasoning among 3-D models and 2-D imagesArtificial Intelligence, 1981
- An interpreter for a language for describing assembliesArtificial Intelligence, 1980