基于GVF模型的地物边界边缘优化方法  

Surface boundary edge optimization based on GVF model

作　　者：彭刚跃 郭珍珍 何齐学 PENG Gangyue;GUO Zhenzhen;HE Qixue(Henan Water&Power Engineering Consulting Company Limited,Zhengzhou Henan 450000,China;Yellow River Hydrological Survey and Mapping Bureau,Zhengzhou Henan 450000,China;Henan D.R.Construction Group,Xinxiang Henan 453000,China)

机构地区：[1]河南省水利勘测设计研究有限公司,河南郑州450000 [2]黄河水文勘察测绘局,河南郑州450000 [3]河南省第二建设集团有限公司,河南新乡453000

出　　处：《北京测绘》2023年第1期100-103,共4页Beijing Surveying and Mapping

摘　　要：采用梯度矢量流(GVF)模型对自动提取的地物边界进行优化,可以在一定程度上提高其与实际边界的匹配度,但GVF模型进行边缘优化在计算过程中容易受到周围突变点或突变区域的影响,导致部分点位偏离真实边界。本文对GVF模型进行改进,分别对道路和海岸线边界进行优化,提高提取结果与实际边界的匹配程度。Gradient vector flow(GVF) model was adopted to optimize the automatically extracted surface boundary, which could improve the matching degree between extraction boundary and the actual boundary to a certain extent. However, the edge optimization of GVF model was easily affected by the surrounding abrupt transition points or abrupt transition regions in the calculation process, which led to the deviation of some points from the real boundary. In this paper, the GVF model was improved by optimizing the road and coastline boundaries respectively to improve the matching degree between the extraction results and the actual boundaries.

关 键 词：边缘优化 梯度矢量流模型 地物提取 

分 类 号：P237[天文地球—摄影测量与遥感]