基于障碍约束的空间聚类算法综述

Survey of Spatial Clustering Algorithm with Obstacle Constrains

传统的空间聚类算法解决的是未带障碍约束的空间数据聚类问题,而现实的地理空间中经常会存在河流、山脉等阻碍物,因此,传统空间聚类算法不适用于带障碍数据约束的现实空间.在解析了带障碍空间聚类相关概念和定义的前提下,对带障碍约束条件的空间聚类算法进行梳理,给出了这类算法的研究历史和沿袭关系,并把这类算法按七个维度分为四大类,分析了每类的技术优缺点,最后给出了带障碍约束的空间聚类算法的未来研究趋向.

Classical algorithms of spatial clustering are performed in optimal data space without any obstacle. But many obstacle constrains exist in the real-world, such as rivers, mountains, etc. They may affect results of clustering substantially. In this paper, the knowledge of spatial clustering algorithm with obstacle constrains is illustrated in brief. And then, research history and inheritance relation of the algorithms is given. These algorithms are divided into four categories from seven respects. At last, technical feature of every category and trend of spatial clustering algorithm in the presence of obstacles are analyzed.