摘要
承灾体是研究地质灾害的一个重要方面,基于高分辨率遥感影像的地物自动提取将成为地质灾害承灾体识别的重要途径。依据面向对象思想,对高分辨率遥感影像进行分割获得承灾体对象;通过对分割后的图像提取图像梯度,进行二值化,去除毛刺等方法处理,获得完整的承灾体边缘。实验结果表明:提取的承灾体边缘完整且准确,在实际应用中具有广泛的适应性,优于传统的提取方法。
It is an important aspect of studying hazard-affected body of geological hazards.Extracting surface feature automatically based on high-resolution remote sensing image will be an important way to distinguish hazard-affected body of geological hazards.On the basis of a method of object-oriented,the hazard-affected body object is formed in the high resolution remote sensing image which is segmented.The complete edge of hazard-affected body object is extracted from the hazard-affected body image which is processed by extracting gradient,binarization,removing spurs,and so on.Experiment results indicate that the hazard-affected body edge is with reasonable completeness and accuracy.The method is of universal application and superior to the traditional methods.
出处
《水文地质工程地质》
CAS
CSCD
北大核心
2011年第1期128-132,共5页
Hydrogeology & Engineering Geology
基金
国土资源大调查工作项目(1212010814037)
关键词
面向对象
承灾体
遥感影像
边缘提取
object-oriented
hazard-affected body
remote sensing image
edge extraction
