像元结构指数的平面道路交叉口提取方法

发布时间：2018-04-06 05:02

本文选题：交叉口提取　切入点：高分辨率　出处：《测绘科学》2017年10期

【摘要】：针对高分辨率遥感影像中道路交叉口难以用固定几何和光谱特征描述的问题,该文提出了一种基于像元结构指数的交叉口提取方法。采用自适应光谱异质性阈值来提取像元形状特征,较好地适应了不同交叉口之间的场景特征差异;构建了像元形状与交叉口结构的量化映射关系,赋予特征语义信息,从而保证了提取结果的合理性。基于城区复杂场景高分遥感影像的交叉口提取实验表明:所提方法能够准确定位交叉口中心点,并检测道路交叉结构;另外,该方法对于交叉口支路发生变化的情况也具有适用性,定性与定量分析表明了该文方法的有效性。
[Abstract]:In order to solve the problem that it is difficult to describe the intersections in high resolution remote sensing images with fixed geometry and spectral features, a method of extracting intersections based on pixel structure index is proposed in this paper.Adaptive spectral heterogeneity threshold is used to extract pixel shape features, which adapts to the difference of scene features between different intersections, constructs the quantitative mapping relationship between pixel shapes and intersection structures, and endows feature semantic information.Thus the reasonableness of the extraction results is ensured.The experiment of intersection extraction based on high score remote sensing image of complex scene in urban area shows that the proposed method can accurately locate the intersection center point and detect the road crossing structure.The method is also applicable to the change of the intersection branch. The qualitative and quantitative analysis shows the effectiveness of this method.
【作者单位】：重庆市勘测院;武汉大学测绘遥感信息工程国家重点实验室;
【基金】：重庆市社会民生科技创新专项(cstc2015shmszx40007) 国家自然科学基金项目(41471354)
【分类号】：P237;TP751

【相似文献】