Time-Geographic Density Estimation for Moving Point Objects

Research output: Contribution to journalArticlepeer-review

Abstract

This research presents a time-geographic method of density estimation for moving point objects. The approach integrates traditional kernel density estimation (KDE) with techniques of time geography to generate a continuous intensity surface that characterises the spatial distribution of a moving object over a fixed time frame. This task is accomplished by computing density estimates as a function of a geo-ellipse generated for each consecutive pair of control points in the object’s space-time path and summing those values at each location in a manner similar to KDE. The main advantages of this approach are: (1) that positive intensities are only assigned to locations within a moving object’s potential path area and (2) that it avoids arbitrary parameter selection as the amount of smoothing is controlled by the object’s maximum potential velocity. The time-geographic density estimation technique is illustrated with a sample dataset, and a discussion of limitations and future work is provided.

Original languageAmerican English
JournalGeographic Information Science
DOIs
StatePublished - Jan 1 2010

Keywords

  • time geography
  • moving objects
  • density estimation
  • point pattern analysis

Cite this