Parallelization implementation of topographic viewpoint filtering algorithm based on terrain viewshed using MPI and OpenMP

The problem of multi-point visibility is an important part in terrain visibility analysis. It is widely used in military, urban planning, protection of endangered animal and other fields. Siting observer point is a kind of multi-point viewshed problem. It is generally abstracted as selecting the least number of viewpoints on a given terrain to maximize the joint viewshed covered by them. A candidate viewpoint filtering method proposed by Wang et al. to effectively solve this problem, includes k-means clustering of candidate viewpoints, calculation of contribution degree of each viewpoint, and ranking of viewpoints on contribution in each cluster. But this method is very time-consuming. Therefore, this paper adopts the MPI parallel program framework to implement the parallelization for the k-means algorithm, and uses OpenMP to realize the parallelization of the ranking process of view contribution of candidate viewpoints and the calculation process of view contribution of each viewpoint. The experimental results show that our parallelization scheme of viewpoint filtering can greatly reduce the calculation time and improve the efficiency of the algorithm.