System for authorizing rendering of objects in three-dimensional spaces
摘要
說(shuō)明書(shū)
SLAM techniques may also include kinematics modeling. Kinematics modeling includes information about action commands given to a robot. As a part of the model, the kinematics of the robot may be included to improve estimates of sensing under conditions of inherent and ambient noise. The dynamic model balances the contributions from various sensors, various partial error models and finally comprises in a sharp virtual depiction as a map with the location and heading of the robot as some cloud of probability. Mapping may be based on the final depiction of the model.
As an example, for 2D robots the kinematics may be given by a mixture of rotation and “move forward” commands, which are implemented with additional motor noise. The distribution formed by independent noise in angular and linear directions may be non-Gaussian, but may be approximated by a Gaussian. An alternative approach is to ignore the kinematic term and read odometry data from robot wheels after each command. Such data may then be treated as one of the sensors rather than as kinematics.
In some cases, SLAM may take into account multiple objects. The related process of data association and computational complexity may involve the identification of multiple confusable landmarks. In some cases, SLAM processes may be posed in terms of multi-object Bayesian filtering with random finite sets that provide superior performance to leading feature-based SLAM algorithms in challenging measurement scenarios with high false alarm rates and high missed detection rates without the need for data association. SLAM may also take into account moving objects.
