摘要
在对NDWI和MNDWI等传统水体指数进行分析的基础上,本文提出了一种创建水体指数的新思路:将多光谱图像进行LBV和HSV等复合变换,综合利用变换前的多波段信息和变换后的新波段信息绘制地物光谱曲线,寻找水体的光谱特性,进行水体指数的创建。本文利用ETM+图像HSV变换后的Sat波段以及LBV变换后的V波段,创建了综合水体指数(SWI),并利用SWI进行了遥感影像水体信息自动提取实验。实验表明:利用SWI能准确提取水体信息,有效减少水体提取结果中的误提取像元。
Traditional water index such as MNDWI and NDWI was analyzed and a new way of establishing water index was pro- posed in the paper. LBV transformation and HSV transformation were carried out on multispectral image. Spectral characteristic curves were drawn and spectrum characteristic of water body was concluded. Water Index was established by using both multispectral informa- tion and new spectral information. In this paper, SWI (Synthetical Water Index) was established by using Sat waveband produced by HSV transformation and V waveband produced by LBV transformation. An automatic extraction experiment of water body information was implemented then. The result showed that SWI could reduce the number of false water pixels (terrain shadow and vegetation).
出处
《测绘科学》
CSCD
北大核心
2013年第4期130-133,共4页
Science of Surveying and Mapping