MPI-Video infrastructure for dynamic environments

Authors:
  • JBJ.E. Boyd
    J.E. Boyd
  • EHE. Hunter
    E. Hunter
  • PKP.H. Kelly
    P.H. Kelly
  • LTLi-Cheng Tai
    Li-Cheng Tai
  • CPC.B. Phillips
    C.B. Phillips
  • RJR.C. Jain
    R.C. Jain

Abstract
We describe an infrastructure for multiple perspective interactive video (MPI-Video) systems. An MPI-Video system assimilates data from multiple disparate sensors, allowing a user to interact with them in a meaningful way. At the heart of an MPI-Video system is an environment model (EM) that both assimilates data from multiple sensors and acts as an information server to clients. A collection of strongly-coupled dynamic state estimation algorithms does the assimilation. Results show that our infrastructure is easy to configure, easy to extend, and is effective at modeling a dynamic environment.