Analysis of surface interpolation algorithms for different surface types
Loading...
Date
2017-01-31
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Digital Terrain Models (DTMs) has become a very significant tool in extracting
geomorphological information from various land areas. Their use is in the fields of mapping,
landscape planning, urban design and many more. Automatic stereocorrelation has been used
to generate a DTM from ASTER stereo image pair (3N and 3B) using ENVI software.
Elevation values were extracted and used with four different interpolation algorithms. The
resulting surfaces when compared with those from the topographic map showed that the
Inverse Distance Weighting (IDW) can achieve a Root Mean Square Error (RMSE) of
±10.773m and Mean Absolute Error (MAE) of ±8.714m for flat terrain as compared to
RMSE of ±11.035m, MAE of ±8.999m for spline; RMSE of ±11.121m, MAE of ±9.102m for
Natural Neighbour (NN) and RMSE of ±12.108m, MAE of ±9.979m for kriging interpolation
method using a point density of 61.49 points per km
2
. Hence, IDW is best for this surface
type. For undulating terrain, IDW again gave the least RMSE of ±13.549m and MAE of
±10.789m in comparison to RMSE of ±13.711m, MAE of ±10.963m for NN; RMSE of
±13.717m, MAE of ±11.028m for spline and RMSE of ±14.835m, MAE of ±11.658m for
kriging interpolation method for point density 62.30 points per km
2
and hence, IDW is again
best for this surface type. For mountainous terrain, NN interpolation method with RMSE of
±19.044m and MAE of ±13.909m gave best results than the other interpolation types. RMSE
of ±21.167m, MAE of ±15.241m was obtained for kriging; RMSE of ±21.632, MAE of
±14.687m for IDW and RMSE of ±21.721m, MAE of ±14.544m for spline for point density
141.64 points per km
2
and so NN works best for mountainous terrains. It is therefore
recommended that IDW interpolation algorithm should be used for both flat and undulating
terrains whereas NN should also be used for mountainous terrains.
Description
A thesis submitted to the Department of Geomatic Engineering, Kwame Nkrumah University of Science and Technology in partial fulfilment of the requirements for the degree of Master of Science in Geomatic Engineering, 2016