The active object recognition stack is a set of components that realize specific services to provide a flexibly applicable object recognition capability for a service robot. It reuses many parts of the mobile manipulation stack. It combines object recognition and manipulation such that out of possible viewpoints of a manipulator-mounted camera the viewpoint with the best information gain for the given task can be determined. It then places the camera accordingly.

This is an overview on the mechanisms of the active object recognition stack.

Explanations to be added:

• Tier 2 services for this stack beyond mobile manipulation services
• Tier 3 Inspection Planning Component

• see the mobile manipulation stack for the baseline reused in this stack
• see D. Stampfer, M. Lutz and C. Schlegel, “Information driven sensor placement for robust active object recognition based on multiple views,” 2012 IEEE International Conference on Technologies for Practical Robot Applications (TePRA), Woburn, MA, 2012, pp. 133-138, doi: 10.1109/TePRA.2012.6215667 for the baseline implementation of the required services and components