Chapter

Unifying Perspectives in Computational and Robot Vision

Volume 8 of the series Lecture Notes in Electrical Engineering pp 11-23

Extracting Planar Kinematic Models Using Interactive Perception

  • Dov KatzAffiliated withComputer Science Department, University of Massachusetts Amherst
  • , Oliver BrockAffiliated withComputer Science Department, University of Massachusetts Amherst

* Final gross prices may vary according to local VAT.

Get Access

Roboticists are working towards the deployment of autonomous mobile manipulators in unstructured and dynamic environments. Adequate autonomy and competency in unstructured environments would open up a variety of important applications for robotics, ranging from planetary exploration to elder care and from the disposal of improvised explosive devices to flexible manufacturing and construction in collaborationwith human experts. Ongoing research efforts seek to enable the use of autonomous robots for these applications through the development of adequate hardware platforms [10, 26, 31], robust and task-oriented control strategies [19], and new learning frameworks [2, 5, 6, 27].

For unstructured and dynamic environments, it is not possible to provide the robot with a detailed a priori model of the world. Consequently, an autonomous robot has to continuously acquire perceptual information to successfully execute mobility and manipulation tasks [12, 17, 25, 29]. This extraction can be performed most effectively, if it occurs in the context of a specific task.