Fast Parameterless Ballistic Launch Point Estimation based on A'-NN Search

摘要

This paper discusses the problem of estimating a ballistic trajectory and the launch point by using a trajectory similarity search in a database. The major difficulty of this problem is that estimation accuracy is guaranteed only when an identical trajectory exists in the trajectory database (TD). Hence, the TD must comprise an unpractically great number of trajectories from various launch points. Authors proposed a simplified trajectory database with a single launch point and a trajectory similarity search algorithm that decomposes trajectory similarity into velocity and position components. These similarities are applied k-NN estimation. Furthermore, they used the iDistance technique to partition the data space of the high-dimensional database for an efficient k-NN search. Authors proved the effectiveness of the proposed algorithm by experiment.