Bayesian fusion of ceiling mounted camera and laser range finder on a mobile robot for people detection and localization

Robust people detection and localization is a prerequisite for many applications where service robots interact with humans. Future robots will not be stand-alone any more but will operate in smart environments that are equipped with sensor systems for context awareness and activity recognition. This paper describes a probabilistic framework for the fusion of data from a laser range finder on a mobile robot and an overhead camera fixed in a domestic environment. The contribution of the framework is that it enables seamless integration with other sensors. For tracking multiple people it is possible to use a probabilistic particle filter tracker. We show that the fusion improves the results of the individual subsystems.