田晶, 何遒, 周梦杰. 运用主成分分析识别道路网中的网格模式[J]. 武汉大学学报 ( 信息科学版), 2013, 38(5): 604-607.
引用本文: 田晶, 何遒, 周梦杰. 运用主成分分析识别道路网中的网格模式[J]. 武汉大学学报 ( 信息科学版), 2013, 38(5): 604-607.
TIAN Jing, HE Qiu, ZHOU Mengjie. Gird Pattern Recognition in Road Network Using rincipal Component Analysis[J]. Geomatics and Information Science of Wuhan University, 2013, 38(5): 604-607.
Citation: TIAN Jing, HE Qiu, ZHOU Mengjie. Gird Pattern Recognition in Road Network Using rincipal Component Analysis[J]. Geomatics and Information Science of Wuhan University, 2013, 38(5): 604-607.

运用主成分分析识别道路网中的网格模式

Gird Pattern Recognition in Road Network Using rincipal Component Analysis

  • 摘要: 提出了基于主成分分析的网格模式识别方法。首先利用形状参量和关系参量描述道路网中的网眼;然后生成网眼数据的主成分;最后,根据主成分分析构造参量权重的思想,导出网眼属于网格模式隶属度,根据该参数识别网格模式。实验结果表明,识别结果与肉眼识别类似。

     

    Abstract: The spatial pattern recognition and implicit information discovery have generated interest in the field of spatial data mining. Grid pattern is one of the most typical patterns in road networks. This paper presents a method for grid pattern recognition using principal component analysis. The method first defines shape measures and relation measures to formalize meshes in road networks. The measures are: rectangularity, convexity, consistent arrangement degree, rectangularity of mesh having largest consistent arrangement degree, convexity of mesh having largest consistent arrangement degree. Second, the principal components are created and analyzed. Finally, the method identifies grid pattern via the meshes’ membership degree of the grid derived from the first principal component. Experiments on Shenzhen road network and Wuhan road network have been conducted. The results show that it is valid in grid pattern recognition and it is a versatile method to some extent.

     

/

返回文章
返回