摘要
区域连接演算(RCC)是定性空间推理的重要基础理论之一。但由于缺乏必要的度量,RCC只是粗略地描述空间拓扑关系而难以对其更准确地描述,也难以利用RCC描述除拓扑关系之外的其它空间关系,如距离、方向等。本文在RCC理论的基础上,提出了区域伸缩演算(RESC)。RESC增加了一个全等CG的原始空间关系,引入了两个新颖的对区域的演算函数,即区域延伸和区域收缩,从而给出了一种以区域为单位的形式化的度量方法。利用RESC,不仅可以扩展RCC-8拓扑关系,而且能以灵活多样的粒度来描述区域间的距离关系、方向关系、位置关系以及运动关系。RESC增强了RCC的空间关系表示能力,拓展了RCC理论的适用范围。
Region connection calculus (RCC) is one of the important fundamental theories in qualitative spatial reasoning (QSR). Lacking necessary metrization, RCC only describe spatial topological relations roughly without further accurateness, and it is also not easy to describe spatial relations such as distance, direction and so on, with the exclusion of topological relation. Based on the RCC theory, Region Extension and Shrinking Calculus (RESC) is proposed. As for RESC, the congruence CG as a primitive spatial relation is added and two new functions acting on region, Region Extension and Region Shrinking are introduced, consequently a formalized metrization method that takes region as a basic unit is put forward. RESC not only expands the topological relations, but also describes multiple spatial relations including distances, directions, positions and locomotion at flexible and various granularities. RESC improves the expressive power of RCC and extends the applicable range of RCC
出处
《计算机科学》
CSCD
北大核心
2008年第4期211-215,共5页
Computer Science
基金
国家自然科学基金项目(60273040)
江苏省高校自然科学研究基金计划项目(03kjd520175)
江苏省社会发展基金项目(BS2001046)
关键词
定性空间表示
区域连接演算
区域伸缩
度量
Spatial relations representation, Region connection calculus, Region extension and shrinking, Metrization