大规模DEM数据并行可视域分析算法研究

Parallel Viewshed Analysis on Large Scale DEM Data

摘要:

可视域分析是具有高时间复杂性的地理计算,在进行大规模数字高程模型(DEM)数据的可视域分析时,传统的串行算法往往难以满足实际需求;提出一种基于环形扫描环的并行可视域分析算法,可以快速计算视点在大范围内的可视域范围,与传统的扫描线算法相比,并行环形扫描线算法以阵面方式向外扩张,从而避免了各坐标点角度的重复计算;通过试验对比发现,并行环形扫描线算法的计算效率约为ArcGIS 10.0中可视域分析的20倍,完全可以满足大规模DEM数据的可视域分析需求。

Abstract:

Viewshed analysis is extremely computing intensive in geocomputation. Moreover, conventional serial algorithm are often difficult to meet this demand. We propose a parallel viewshed algorithm based on scanning circle which can compute the viewshed area quickly on large scale Digital Elevation Model (DEM). Compared with conventional algorithms, our algorithm computes the viewshed area in an expanded way which can help to avoid the redundant computation. Experiments demonstrate that our algorithm is 20 times faster than the viewshed analysis in ArcGIS 10.0, and can meet the demand of viewshed analysis on large scale DEMs.