Hierarchical spatial relation based on a contiguity graph

Abstract

Autonomous agents traversing a natural space need to be knowledgeable of its space configuration. The existing space models in geographic information systems and robotics, however, deal with only the topological relations among numerous types of spatial relations. We aim to develop an enhanced space model that elaborates the spatial relations with respect to their relevant physical relations. Specifically, the spatial relations in a space configuration are further characterized with the gravitation as a potential factor to affect the space configuration. The resulting space model is capable of capturing an extended set of spatial relations over existing models such as a four-intersection model.

A Method of Extracting Inter-Object Spatial Relationship in Remote Sensing Image Objects

“…As a crucial spatial relation, topological relation has the features of scale, cognition and hierarchy as well as other spatial. In general, the hierarchical feature of spatial relation is described by 2D-String, the quadtree, K-dtree, Btree, Rtree, and so on [6,7]. In these hierarchical tree structures, the root of tree is on behalf of the highest level of spatial objects; the leaves of tree is on behalf of the lowest level of atomic objects.…”

Spatial Topological Relationship for Remote Sensing Image Analysis

“…Cognitive feature: Topological relationship is a direct reflection from human cognitive concept of the real world to identified images [8]. Remote sensing image analysis and understanding must meet the human needs to understand and express the real environment.…”

Fuzzy Spatio-temporal Relations Analysis

“…This method is inspired by Max J. Egenhofer's method [2] of topological distances by considering the directional relations. In situations like in figure 3, where the distance and directional information becomes invalid and the object does not change the topological relations Sung-Hye et al [10] introduced the method of hierarchical spatial relations based on contiguity graph. This method depends upon the physical properties of objects not on the spatial properties.…”