一种基于拱高半径复变函数的面实体匹配算法

Polygon entity matching algorithm based on arc-height radius complex function

针对面实体匹配问题进行了研究。面实体的边界线在某点的拱高正是对边界线在该点的弯曲程度和凸凹性的反映,该点的中心距离又可以对面实体形状的整体进行描述,通过边界线上某点的中心距离和拱高组成复数,并对其进行快速傅里叶变换可以获取傅里叶形状描述子,作为对面实体形状相似度的度量。将面实体的空间位置、形状、大小等相似度通过加权综合,获得了一种综合空间相似度度量模型,利用此模型对面实体进行匹配。实验结果表明,算法能够有效地进行面实体的匹配。

Study on the problem of polygon entity matching.The arc-height at some point on the boundary line of a polygon entity is the degree of boundary line bending at that point and is the convexity and concavity reflection of the point,the center distance of the point can describe the overall shape of the entity,so the center distance as real part and the arch as imaginary part compose a complex,then its fast Fourier transform obtain the Fourier shape descriptors,as a measure of shape similarity.This paper proposed a composite space similarity model by integrating location similarity,shape similarity and size similarity with a weighted composite method,and used the model to match polygon entities.Experimental results show that the algorithm can effectively carry out the matching polygon entities