PhD forum: Competing agents for distributed object-tracking in smart camera networks

Abstract

This paper describes an approach for object-tracking in large smart camera networks, which is based on software-agents with conflicting goals. The focus is on the system architecture. These software-agents are specialized on concrete functions, e.g. maximizing sensor coverage of a surveillance area by aligning the cameras' fields of view. To reach collaborative behavior, an approach inspired by Multi-Criteria Decision Making (MCDM) of the operations research is used.