STUDY AND APPLICATION OF PARALLEL COMPUTING BASED-ON MPI IN SITUATION ASSESSMENT SYSTEM FOR DIGITAL ARMY

摘要

Situation assessment system for digital army is the important part of the whole command information system. This paper deals with the parallel computing in the object recognition of situation assessment system, analyses its feasibility, provides a strategy of partitioning based on track and a method to allow the Dempster-Shafer theory of evidence work in parallel, designs and implements the algorithm of parallel object recognition with the Message-Passing Interface based on a dynamic load balancing technology. Tested by series of experiments, this parallel algorithm based-on MPI improves the speed of recognition while ensuring the reliability of object recognition for situation assessment system.