一种基于种子点区域增长的简单地物边界快速提取方法  

A Fast Edge Extraction Method of Simply Objects Based on Seed Region Growing

作　　者：左琛[1] 

机构地区：[1]北京市测绘设计研究院,北京100038

出　　处：《城市勘测》2017年第5期126-129,共4页Urban Geotechnical Investigation & Surveying

摘　　要：从遥感图像上提取地物地貌信息,并对其进行处理和分析,广泛应用于城市规划、变化监测、环境保护、应急减灾等各个领域。本文针对测绘规划部门常用的真彩色遥感影像图,综合国内外相关研究基础上,提出一种基于种子点区域增长算法快速提取简单地物边界的方法。并以影像图中水体、云斑为实验对象验证方法的有效性。Extracting and analyzing ground object information from remote sensing images is widely applied in many areas,such as urban planning,land usage changing monitoring,environment protection and disaster relief.For true color remote sensing images commonly used by survey planning department,this paper synthesized domestic and international related research and proposed a fast edge extraction method,which is based on seed region growing algorithm.Finally,this paper verified this method by experiments on water and cloud in remote sensing images.

关 键 词：图像分割 种子点区域增长 阈值 边缘提取 

分 类 号：P283.8[天文地球—地图制图学与地理信息工程] TP753[天文地球—测绘科学与技术]