基于高程分层方法的HJ-1B CCD2影像大气校正

大气校正是从环境一号B星(HJ|1B)CCD2多光谱卫星数据中精确提取地面定量信息的关键一步。采用一种基于高程分层和改进的浓密植被算法,即将整个研究区按照高程间隔0.1 km划分为17个子区域,在不同高程带内利用红波段与短波红外波段(1.6 μm)的线性关系估计出红波段的反射率,然后利用估计的红波段反射率与其表观反射率差值的均值,结合6S辐射传输模型模拟计算得到0.55 μm处的气溶胶光学厚度,从而实现各个波段大气校正。比较分析校正前后成熟林地、水体和裸土的光谱反射率和标准反射率,结果表明:该方法对HJ-1B CCD2数据大气校正可取得较好的效果。

Atmospheric Correction for Environment Satellite CCD2 Data based on Stratified Elevation

To extract quantitative information from HJ-1B imagery data accurately,atmospheric correction is a necessary step.We present a new algorithm that based on stratified elevation and using the spectral correlation between red and 1.6 μm channel to drive atmospheric properties and execute atmospheric correction process for HJ-1B Data.Since our study area in mountainous region,we separate the whole area into 17 different elevation zones by step of 0.1 km and simulating aerosol optical depth in 0.55 μm within each elevation zone using radiative transfer model,then the atmospheric correction procedure in different elevation zone of each visible,infrared and short wave infrared spectral channel is performed.Through the comparative analysis of spectral curve of dense mature forest,bare soil and water contained sediment before and after atmospheric correction at different sites,the result indicating this approach is promising in atmospheric correction for HJ|1B imagery data.