定性空间推理(QSR)研究空间关系,多数工作集中在单维空间关系,但在地理信息系统(GIS)中多维对象很常见.混合维对象空间关系是指点、线和区域3类对象出现在同一场景的情况,该类问题对定性空间推理研究有着重要的理论意义和应用价值,但这方面的研究工作还比较少.在已有的混合维区域连接演算的基础上进行完善,提出了MRCC5混合维拓扑模型,并研究了其上约束满足推理问题的复杂度.对定性尺寸关系进行了混合维扩展,给出了MDS模型,进而研究了其推理问题.在以上工作基础上,提出了RCC5和MDS的结合模型,给出并分析了结合模型的推理算法.将定性空间推理相关研究推广到混合雏领域,深入研究了混合雏拓扑关系推理,提出了混合维尺寸以及混合雏拓扑尺寸结合模型.
Qualitative spatial reasoning (QSR) studies the spatial relations of objects in the space. Most QSR research work aims at the spatial relations between single dimensional objects (i. e. region- region relation or line-line relation). Spatial relation of multi-dimensional objects is the situation that point, line and region objects appear in one scene at the same time. Spatial relation of multi- dimensional objects is very common in geographical information systems (GIS). Qualitative spatial relation of multi-dimensional objects has very important theoretical meaning and practical value, but very few work in QSR was focused on it before. By improving the previous research on multi- dimensional region connection calculus model, a multi-dimensional topological model (i. e. MRCCS) is given, and the complexity of its constraints satisfaction problem is studied. The multi-dimensional extension of qualitative size relation called MDS model is proposed, its reasoning problem is also investigated. Based on the above work, a model which combines RCC5 and MDS is proposed, so does its reasoning algorithm. This work extends the related work in qualitative spatial reasoning to the multi-dimensional field. The reasoning of multi-dimensional topological relation is improved. It is the first time that multi-dimensional size relation model and relation model integrating topology and size are studied.