摘要
以江苏苏北地区为研究区,较系统地分析了目前世界上星载传感器中光谱分辨率最高的EO-1卫星Hyperion数据的基本特点和处理方法,研究了将图像的灰度值转变为辐射能量值和反射率的方法和步骤,并选取了15类典型地物,进行了样本采集、辐射亮度和反射率反演,最后对典型地物的光谱曲线进行综合分析研究,并概括了苏北沿海滩涂典型地物的基本光谱特征。
Hyperspectral remote sensing technology is one of the most important leading fields in remote sensing which is paid more and more attention in the world. The successful launch of Hyperion in November 2000 provided a new and useful tool for remote sensing applications. Hyperion imaging spectrometer has the highest spectral resolution in orbiting satellites, and shows great application potential. Taking Subei district, Jiangsu province as a study area, this article analyzed the characteristic of EO-1 Hyperion data, studied the general method and procedure of calculating the spectral value of image from its grey value, selected 15 kinds of typical object samples and deduced their spectral curves in wavelength ranging from 400 nm to 2 500 nm. The feature of objects can be seen clearly from deduced spectral curves. Since the imaging time of the data was in late winter (10:19 am, January 15, 2003), the vegetation withered and no typical spectrum characteristics of vegetation were displayed in visible wavelength. But in near-infrared region there was still a high reflecting peak. In short wave infrared wavelength, the reflection value is lower than that of near infrared because of the influence of water absorption. At the high tide time, the difference between river water and seawater was evident. The reflectance of seawater was much higher than that of river water because the seawater contained more mud and sand. The reflection value of turbid water is lower than that of clean river water in green wavelength and higher in red and infrared wavelength. Probably this is caused by the higher content of organic material and algous vegetation in turbid water. In salt field, the spectral reflection values of depositing pool, evaporation pool and crystal pool rose in turn with the increase of salt amount.
出处
《地球信息科学》
CSCD
2004年第2期81-87,共7页
Geo-information Science
