Environment Models and Sensor Integration
The capabilities of a robot system are strongly dependent on an ability to observe and understand the environment in which it operates. For this reason, an effective and purposeful description of the environment is fundamental to the development of intelligent robotics [Ambler 86]. There are many different ways in which the problem of description can be approached, resulting in a variety of different environment modeling techniques; solid models, primitive surface models, or functional descriptions, for example. The utility of each description depends on the specific task in which it will be used. Ideally a robot system should provide many different levels of environment description, and be able to move freely and maintain consistency between them.
KeywordsManifold Covariance Propa Eter Cross Correlation
Unable to display preview. Download preview PDF.