摘要
建筑物作为地理信息基础数据,是衡量城市发展的主要指标,如何对遥感影像对建筑物进行的提取是遥感图像处理的热点。本文研究了基于面向对象的高分遥感影像建筑物提取,首先对影像进行多尺度分割,然后对分割以后形成的有意义的图斑进行处理。结合建筑物的光谱、形状等特征对建筑物进行提取,实验结果表明该方法提取结果较好,精度可以达到90.3%。
As the basic data of geographic information,buildings are the main indicators to measure urban development.How to extract buildings from remote sensing images is a hotspot of remote sensing image processing.In this paper,we study the object-based high-resolution remote sensing image building extraction.Firstly,the image is multiscale segmented,and then the meaningful maps formed after segmentation are processed.The building is extracted according to the characteristics of the building’s spectrum and shape.The experimental results show that the method has good extraction results and the accuracy can reach 90.3%.
作者
吕道双
林娜
张小青
Lü Daoshuang;LIN Na;ZHANG Xiaoqing(College of Civil Engineering,Chongqing Jiaotong University,Chongqing 400074,China)
出处
《北京测绘》
2019年第2期191-195,共5页
Beijing Surveying and Mapping
基金
重庆市教委科学技术研究项目(KJQN201800747)
重庆市留学人员回国创新支持计划项目(CX2017127)
关键词
高分辨率遥感影像
面向对象
多尺度分割
建筑物提取
high resolution remote sensing image
object-oriented
multi-scale segmentation
building extraction
