Feature Extraction And Reconstruction Methods Research Of Complex Surface

Surface reconstruction placed great emphasis on digital terrain analysis, fitting reconstruction and interpolation reconstruction are widely studied surface reconstruction theory, however, both of them have defects like intensive computing capacity, terrain skeleton points cannot be determined to be retained, weak control to terrain noise and so on, especially for complex surface's reconstruction herein. In this paper, the complex surface is reconstructed based on the idea of Fourth Order Partial Differential under valley lines and ridge lines constraints, a good solution to the above problem.As the characteristics constraint of surface reconstruction, extracting valley lines and ridge lines is the focus of this paper, but the traditional flow simulation algorithm presences algorithm complexity and time-consuming shortcomings when extracting valley lines and ridge lines, unable to meet the large-scale, irregular terrain point feature extraction requirements of this study. To solve this problem, on the basis of traditional water simulation algorithm, this essay designs triangulation-based feature extraction algorithm, and the algorithm greatly improves the efficiency of valley lines and ridge lines extraction by full use of the hash thought. This paper's contribution are as follows:1. Researched the way to triangulate the large-scale, irregular distribute input surfaces point set, and based on the generated triangulation which in line with the DELAUNAY split guidelines, improved the traditional water simulation algorithm and designed the hash thought based algorithm to extract the valley lines and ridge lines. The test verified that the designed feature extraction algorithm meet the complexity requirements and less time-consuming.2. Researched the way of complex surface reconstruction and designed the Gaussian filter method based on triangulation data, it was proved that the terrain noise was greatly suppressed after filtered. Additionally, this article has improved surface reconstruction algorithm based on Fourth Order Partial Differential Equations, by introducing characterized constraints to preserve topographic features in the process of surface smooth.3. Implemented the feature extraction and complex surface reconstruction system, the system integrated the main functions like original surface points data's import and display, triangulation and interpolation of points data, extraction of surface feature lines and points(mainly for the valley lines and ridge lines), surface reconstruction based on triangular net, gridded surface data and Fourth Order Partial Differential smoothing.By contrast with the features extraction and complex surface reconstruction result from ArcGIS, a mainstream terrain processing software in market, it was proved that the feature extraction and complex surface reconstruction method herein is feasible and efficient, which provides a new way of the thinking to the feature extraction and surface reconstruction theory based on large-scale and irregular terrain surface points.