Succinct quadtrees for road data

Abstract

We propose succinct quadtrees, space-efficient data structures for nearest point and segment queries in 2D space. We can compress both the tree structure and point coordinates and support fast queries. One important application is so called map matching, given GPS location data with errors, to correct errors by finding the nearest road. Experimental results show that our new data structure uses 1/25 working memory of a standard library for nearest point queries.