Abstract
Mobile tracking technologies are facilitating the collection of increasingly large and detailed data sets on object movement. Movement data are collected by recording an object’s location at discrete time intervals. Often, of interest is to estimate the unknown position of the object at unrecorded time points to increase the temporal resolution of the data, to correct erroneous or missing data points, or to match the recorded times between multiple data sets. Estimating an object’s unknown location between known locations is termed path interpolation. This paper introduces a new method for path interpolation termed kinematic interpolation. Kinematic interpolation incorporates object kinematics (i.e. velocity and acceleration) into the interpolation process. Six empirical data sets (two types of correlated random walks, caribou, cyclist, hurricane and athlete tracking data) are used to compare kinematic interpolation to other interpolation algorithms. Results showed kinematic interpolation to be a suitable interpolation method with fast-moving objects (e.g. the cyclist, hurricane and athlete tracking data), while other algorithms performed best with the correlated random walk and caribou data. Several issues associated with path interpolation tasks are discussed along with potential applications where kinematic interpolation can be useful. Finally, code for performing path interpolation is provided (for each method compared within) using the statistical software R.
Acknowledgements
The author wishes to thank U. Demšar for comments on a previous draft of this manuscript. The author also thanks the three anonymous referees and the Associate Editor S. Dodge who provided critical feedback that greatly improved this work. The caribou telemetry data were made available by the British Columbia Ministry of Environment. The athlete GPS data were collected in collaboration with the University of Victoria Ultimate Frisbee Club.
Disclosure statement
No potential conflict of interest was reported by the author.
Supplemental data
Supplemental data for this article can be accessed here.