Search

In the integrated generalization of contours and rivers, it is more important to consider the generalization of their geographical characteristics in addition to focusing on the graph generalization. Therefore, we elaborated on the geographical characteristics on which the contours and rivers depend in the integrated generalization, qualitatively describe and quantitatively characterize the geographical characteristics (spatial characteristics and geographical elements information content characteristics), and put forward the integrated generalization method of contours and rivers considering geographical characteristics. Based on the geographical characteristics, this article first identifies the relief and river network morphology, then as an essential terrain constraint, the spatial analysis and selection of topographic feature lines were carried out. Second, according to the integrated generalization method, the contours and rivers were discretized into 3D points along with describing semantic information and giving weights. Finally, the integrated generalization considering geographical characteristics experiment was implemented and well generalization results were obtained.

The existing integrated generalization method for contours and rivers using a 3D Douglas-Peucker algorithm is somewhat inefficient because it reserves points by constantly setting thresholds for exploring which one is suitable for the target scale. Here, we present an improved integrated generalization method for contours and rivers by applying an importance sequence scheme upon all points from these two elements. Our key method is an improvement upon the point selection using the 3D Douglas- Peucker algorithm. At first, no points were deleted, and the importance of all points from contours and rivers with geometric and semantic weights were calculated. Then, an importance sequence was formed which ordered all points from the largest to the smallest importance values. Finally, we adjusted the location of quantiles to reserve the necessary points for target scales. Results show that the proposed method can avoid the spatial and semantic conflicts between contours and rivers, and improve the cartographic generalization efficiency. [ABSTRACT FROM AUTHOR]