Interpolation methods for high-fidelity three-dimensional terrain surfaces

Recent developments in high-fidelity three-dimensional terrain surface measurements have necessitated the development of interpolation methods that are specific to different ground vehicle applications. The objective of this work is the development of application-dependent methods to convert non-uniformly spaced data to uniformly spaced data. This work develops techniques for applying interpolation methods to on-road and off-road terrain surfaces that are used in vehicle simulations. The interpolation methods examined to create uniformly spaced terrain surfaces are mean, median, Inverse Distance to a Power, and Kriging. Each method is judged in four areas: computational efficiency, outlier sensitivity, location sensitivity, and trend sensitivity. The results of this work show that for applications where the resolution of interest is coarse (25 mm or greater) the mean and median are advantageous for creating a uniform grid; but for data applications where fine resolution is required (less than 25 mm) the Inverse Distance to a Power and Kriging have advantages. (C) 2010 ISTVS. Published by Elsevier Ltd. All rights reserved.