地标空间方向的位置描述定位模型

A positioning localities model about landmark locality description based on spatial direction relationship

针对地标系统中方向位置描述中位置线索少而难以实现定位的问题,该文首先对位置描述的语义上下文进行分析,得到人们在用方位关系进行位置描述时,会潜在地选择位置"附近"的地标作为参考点;然后,从人们对方位的认知与角度相关出发,建立方位关系方位隶属度函数,基于欧式距离与均摊面积(Stolen-Area)建立"附近"的隶属度函数,利用联合概率描述目标位置实现定位,同时给出了描述定位精度的误差参数;最后,通过算例对模型进行验证。实验结果表明,该方法能够显著提高定位精度。

Aiming at solving the problem that there were few position clues in the directional locality description in the landmark system and it was difficult to achieve positioning,this paper analysed the semantic context of locality description and found that the near landmarks were selected easily when describing localities. To achieve positioning,a joined probability function which consisted of distance and direction membership functions based on spatial cognitive was proposed. Based on human’s direction cognition was related to the angle,the direction membership function was provided. The "near" relation membership function was conducted based on the Euclidean distance and stolen area. Meanwhile,the error parameters for positioning uncertainty were defined. Finally,same examples were given to demonstrate our model. Experimental results showed that this method could significantly improve the positioning accuracy.