A computationally efficient and cost effective multisensor data fusion algorithm for the United States Coast Guard Vessel Traffic Services system

Abstract

This thesis develops an algorithm to fuse redundant observations due to multiple sensor (type and location) coverage in order to provide a significant reduction in duplicate track information provided to Vessel Traffic Services (VTS) operator displays. The design of the algorithm allows acceptance of inputs from any type of sensor (radar, acoustic, OPS, system generated and manual tracks) as long as the basic decision criteria elements are provided. The result of this effort is a computationally efficient and cost effective software solution to a significant system deficiency that impacts greatly on overall waterway safety. The algorithm is tested with real data collected from the VTS system at Puget Sound in September 1996. The results indicate that the algorithm correctly fuses redundant sensor observations on the same vessel resulting in a significant reduction in the amount of unnecessary information presented to the VTS operator.