An efficient algorithm for the extraction of topographic structures from large scale grid DEMs

As a fundamental problem in terrain analysis, the extraction of topographic structures plays a role in many applications such as hydrographic analysis, mineral deposition, land erosion pollution diffusion analysis, and so forth. The inherent problems in the efficiency and accuracy of existing approaches to extraction have hindered their application in the processing of large-scale Digital Elevation Models (DEMs). In this paper, an efficient approach to the extraction of topographic structures based on grid DEMs is proposed. The authors make two significant improvements to the extraction of topographic structures. The first is to fill depressions by adopting vector processes combined with traditional neighbourhood raster processes, and the second is to assign flow direction over flat by applying a neighbour-grouping scan method. After the experimental analysis using various volumes of real DEMs, the results show a significantly improved level of efficiency in the extraction of topographic structures, especially when applied to large-scale DEMs. The accuracy of the results is also better than existing methods. Thus, these improvements make the digital terrain analysis method more practical for large regional spatial decision-making.