基于自适应采样粒度模型的空间方向关系模糊描述方法

A Fuzzy Description Method for Directional Relationship Based on Adaptive Sampling Granularity Model

针对空间方向关系具有不确定性的特点以及现有的方向关系表达方法的局限性,结合模糊理论分析方向隶属函数的基本特征,建立基于非线性隶属函数的8方向模糊描述参考框架。分析空间目标间的距离、尺寸、形状等因素对方向关系的影响,提出空间目标的自适应采样粒度模型。在保证空间方向关系一致性的前提下,对空间对象进行采样处理生成合适大小的点集。在8方向模糊描述参考框架下,统计点集之间的8个方向模糊矩阵,得到空间对象之间在各方向概念上的隶属程度,实现空间方向关系的细节模糊描述。通过实例分析验证该模糊描述方法的可行性和有效性。

Due to the vagueness of spatial directional relationship, the basic essential characteristics of fuzzy directional membership functions are discussed. And the reference framework of 8-directions fuzzy description based on nonlinear functions is established, which is integrated with fuzzy theory and existing experimental results. Then, by analyzing the influence of factors of directional relationship such as the distances between objects, and the shapes and sizes of objects, an adaptive sampling model is presented, which can transform the directional relationship of spatial object into the sampling point sets. Therefore, 8 directional fuzzy matrixes are obtained with this model to do a quantitative calculation. Through the statistics of these fuzzy matrixes, the detailed fuzzy description of directional relationship is achieved. At last, a concrete instance and the result of experiment show that not only the appropriate sampling granularities were chosen automatically, but also the detailed fuzzy description of directional relationship between arbitrary spatial objects can be achieved effectively using this model.