Exploring the effect of Steiner points on the simplification algorithms

Abstract

With the increasing volume of spatial data generated by a variety of spatial data recording tools such as smartphones, the importance of geometric simplification approaches has become more and more over time. The goal of geometric simplification is to achieve more summarized and less complex features. It provides an algorithm that results in terms of geometric properties such as area, perimeter, and angles being more similar to the primary feature. Algorithms with lower accuracy select consecutive subsets of primary points. As a result, some points of the geometric shape are completely ignored. While the results of methods such as least squares (LS) are more accurate in geometric simplification. Also, most geometric simplification algorithms of linear features focus on points in their processes and ignore the edges. Therefore, in this study, to improve the accuracy of geometric simplification accuracy, the effect of Steiner points on Douglas Poker (DP), LS, and a combination of them (DP-LS) was investigated. For this purpose, the trajectory recorded in Einali Mountain of Tabriz was used. The results showed that the use of Steiner points on average led to an improvement of 3.46% angle changes, 914941 m2 area difference, 2.66% curvature similarity, and 0.36% node reduction in DP-LS and LS methods.