An improved simulated annealing algorithm for road stroke construction in Taibei
Stroke is a set of ordered sections in the same direction,and it is also a high-level semantic of the road network,widely used in road network mapping synthesis.It is necessary to follow the principle of'good continuation'when constructing strokes.The primary aim of this study is to address the challenges inherent in traditional methods of stroke construction within road networks,extensively used in road network cartographic generalization.Traditional approaches often rely on setting various thresholds for pairing road segments,a task complicated by the inability to determine optimal thresholds in advance,sensitivity to coordinate errors,and the lack of multi-scale information.To overcome these issues,the study proposes a novel method for stroke construction based on an objective function and global optimization,emphasizing the principle of'good continuation'.The innovative methodology proposed in this study emphasizes the integration of local smoothness and global extendibility criteria,balanced through a carefully chosen weighting factor.The research employs a case study approach using vector road data from Taibei,comprising both points,called nodes in programing,and lines,called edges in programing,which undergo vector boundary clipping as a preprocessing step.Through graph theory modeling,the point and line datasets are utilized to construct a network of nodes and their adjacency relationships.The core of the method involves the application of a simulated annealing algorithm for the global optimization of stroke construction.This new approach is rigorously compared with traditional methods to assess its efficacy and efficiency in improving road network depiction.The application of this method to the Taibei vector road network dataset reveals significant improvements over conventional techniques.Specifically,the proposed method successfully increases the average length of strokes and reduces the number of strokes required to represent the road network,indicating a more streamlined and coherent map.Remarkably,the stroke sets constructed using the proposed methodology exhibit a 15.9%increase in accuracy compared to those produced by traditional methods.This indicates a substantial enhancement in the ability to clearly delineate the principal routes and hierarchical organization of the road network,facilitating better navigation and understanding.The result set of strokes clearly recognizes the backbone and hierarchical relationships of the road network.The research concludes that the proposed method for stroke construction,leveraging an objective function and global optimization,significantly enhances the clarity and coherence of road network representations.By incorporating local smoothness and global extendibility criteria,and employing simulated annealing for optimization,the method effectively addresses the limitations of traditional approaches.The successful application to the Taibei road network dataset highlights its practical utility and potential for broader application in cartographic generalization.The potential applicability of the method to different geographical environments is also demonstrated.This innovative approach facilitates the creation of accurate and intuitive road network maps in various geographical environments.The result set of strokes constructed by this method is helpful for the subsequent research of road intersection recognition.
multiscale judgment criteriaroad modelingstroke buildingglobal optimizationsimulated annealinggraph theory
NETL
NSTL
万方数据
