Determining the relationship between the slope and directional distribution of the UAV point cloud and the accuracy of various IDW interpolation

Inverse Distance Weighted (IDW) based interpolation method is also widely used in earth science studies. In the classical IDW method, the directional distribution of the reference points around the point to be estimated within the critical circle and the slope differences are not taken into consideration. On the other hand, in the IDW-based method developed by Shepard, the ratio of the distances of the reference points within the critical circle to the critical circle radius (r), the anisotropy and the slope differences are taken into consideration. In this study, the results of the classical IDW method and Shepard method were compared to increase the accuracy of interpolation produced from UAV data. A software has been developed to make these comparisons in more detail. The classical IDW and Shepard based interpolation methods used in this software takes into consideration the anisotropy, the slope differences and the ratio of the distances to the critical circle radius. In this study, UAV flights were performed in three different study areas with different topographic features and 3D point cloud data were obtained in order to make detailed analyzes. Using developed software, data from three different study areas have been tested and the results from different Shepard interpolation models have been discussed. The major contribution of this paper is in evaluation of various IDW options when applied to UAV point data. As a result, especially in geodetic studies form UAV data, it was observed that the results improved with 11% to 37% by using the Shepard method with the suitable power parameter value considering the directional distribution of the reference points in the critical circle and the slope differences.