摘要
已有的拓扑相似性度量方法主要通过目标之间拓扑关系统计均值的差值,衡量目标之间的拓扑相似程度,在拓扑变化复杂区域由于拓扑关系的增减互补易产生度量误差。为此,提出一种顾及拓扑结构差异的等高线群拓扑相似性度量方法;引入Zhang-shasha算法将等高线群之间的拓扑变化转化为等高线树之间的转换编辑操作;构建基于树结构的拓扑差异表达;并利用动态规划方法求解等高线树之间的转换编辑距离;进而量化度量等高线群间的拓扑结构相似度。模拟数据实验和真实数据实验表明,该方法顾及了拓扑结构差异,能够有效度量多源多尺度等高线数据之间的拓扑相似程度,减少度量误差。
In previous topological similarity measure methods,the difference between statistical mean values are mainly calculated for the topological similarity and the topological structure variation are not considered,which may lead to measurement errors in areas with complex topological changes.Therefore,a topological similarity measure method is proposed based on topological structure variations.Based on Zhang-shasha algorithm,the topological changes between the contour groups are converted into the editing operation between the contour trees.Topology difference expression based on tree structure is constructed and the transformation distance between the contour trees is calculated through dynamic programming method.Then the topological similarity is quantitative measured based on transformation distance.Simulated and real datasets experiments prove the proposed measure method can effectively measure the topological similarity between multisource contours and reduce the measurement error.
作者
郭文月
刘海砚
孙群
余岸竹
季晓林
GUO Wenyue;LIU Haiyan;SUN Qun;YU Anzhu;JI Xiaolin(Information Engineering University,Zhengzhou 450001,China)
出处
《测绘科学技术学报》
北大核心
2019年第1期79-85,共7页
Journal of Geomatics Science and Technology
基金
国家自然科学基金项目(41801388
41571399
41801313)
