Evaluation of the performance of DEM interpolation algorithms for LiDAR data

This is the latest version of this eprint.

Abstract

Airborne light detection and ranging (LiDAR) is one of the most effective means for high quality terrain data acquisition. The high-accuracy and high-density LiDAR data makes it possible to model terrain surface in more detail. Using LiDAR data for DEM generation is becoming a standard practice in the spatial science community. Of the three commonly used digital elevation models (e.g., triangular irregular network (TIN), gridded DEM and contour line model), the gridded DEM is the simplest and the most efficient approach in terms of storage and manipulation. However, this approach is liable to introduce errors because of its discontinuous representation of the terrain surface based on the interpolation process of sampled terrain points. Given the characteristics of LiDAR data, much attention must be paid to the selection of an appropriate interpolation algorithm, otherwise the accuracy of produced DEM from LiDAR data will be compromised. This study aims to evaluate the performance of commonly used interpolation algorithms to the LiDAR data, including inverse distance weighted (IDW) method, Kriging method, and local polynomial method. All these interpolation algorithms are applied to DEMs generated from LiDAR at various data density levels. The performance of these interpolation methods is evaluated by using both cross-validation and validation test methods. The results showed the performance of each interpolation algorithm for two study sites with different terrain types and analysed the relationship between interpolation algorithms and LiDAR data density. Considering accuracy and computing time for large volume of LiDAR data, IDW is recommended for LiDAR DEM generation from this study.

Available Versions of this Item

• Evaluation of the performance of DEM interpolation algorithms for LiDAR data. (deposited 06 Jan 2010 10:04) [Currently Displayed]