Efficient data structures for large scale tracking

This paper describes a set of data structures that enables the tracking of large scale data sets. Although well-known procedures exist for speeding up tracking performance, such as gating, they are typically not sufficient for situations where it is required to simultaneously track tens or hundreds of thousands of targets where even the gating calculations themselves take a significant proportion of time. We describe dynamic spatiotemporal binary tree-based structures, a box forest and cone forest, for storing measurements and tracks, and a string trie for text information. Efficient pruning of the structures allows for a vast reduction in the number of gating calculations. Performance of a real-time tracking algorithm that uses the data structures is demonstrated on a real-world maritime data set of more than 100,000 targets.