Fusion architectures for intelligent integrated processing of multisensor data in target surveillance and tracking applications

Multisensor operations in target surveillance and tracking applications demand implementation of efficient fusion architectures that support intelligent integrated processing of incoming data streams. A prime requirement in these implementations is the exploitation of the synergy available in the multiple representations of the scene under surveillance obtained from different sensor measurements without incurring any major increases in computational complexity. A generic architecture that performs a feature-level fusion of measurements from diverse sensors used in surveillance and tracking environments is presented. Some illustrative designs of fusion architectures for tracking target maneuvers and for superresolution of imagery and image sequences are also presented.