地理空间信息2024，Vol.22Issue(5)：53-56,65,5.
遵循格式塔原则的多约束地图居民地聚类研究
Aggregation Algorithm of Multi-constraint Map Residential Areas Based on Gestalt Principle
摘要
Abstract
Taking the residential areas aggregation on map as the research object,from the perspective of constraint conditions and clustering algorithm,combining with the Gestalt principle of residential distribution mode,we used the minimum spanning tree(MST)based on the constrained Delaunay triangulation to optimize the"visual distance"proposed by AI Tinghua.In order to provide reference for automatic cartographic synthesis of map elements,we replaced Euclidean distance with neighborhood distance,added shape and density differences,and added semantic constraints of building height.By comparing the clustering results of MST under Euclidean distance,proximity distance and visual distance and the clustering results of K-means method,systematic clustering method and MST algorithm under visual distance constraints,we found that the MST clustering algorithm based on the optimized visual distance was more effective and more consistent with human cognitive result.关键词
制图综合/居民地聚类/视觉距离/格式塔原则/MSTKey words
mapping generalization/residential areas aggregation/visual distance/Gestalt principle/MST分类
天文与地球科学引用本文复制引用
蔡娇楠,杨德芳,孟妮娜..遵循格式塔原则的多约束地图居民地聚类研究[J].地理空间信息,2024,22(5):53-56,65,5.基金项目
国家自然科学基金资助项目(41501498) （41501498）
陕西省自然科学基础研究计划资助项目(2021JM-155) （2021JM-155）
高分专项黄河流域青海段生态保护和高质量发展应用产业化示范项目. （）
