CAJ | 학술논문

遥感影像的大气校正是遥感数据应用的基础．文章利用ENVI软件的FLAASH大气校正模块，对Hype—rion高光谱遥感影像进行大气校正，选取典型地物对校正效果进行评价．主观定性评价结果显示校正后的影像地物边界更清楚、对比度增强；光谱曲线对比分析显示，校正前后光谱曲线形状完全不同，林地的反射率图像的光谱曲线形状与USGS标准库的曲线类似；典型地物影像质量的指标评价结果表明，校正后图像的平均梯度值基本都大于校正前影像的值；信息熵的值正好相反，大气校正的效果明显，进一步证实FLAASH模块各参数设置的合理性．
요감영상적대기교정시요감수거응용적기출．문장이용ENVI연건적FLAASH대기교정모괴，대Hype—rion고광보요감영상진행대기교정，선취전형지물대교정효과진행평개．주관정성평개결과현시교정후적영상지물변계경청초、대비도증강；광보곡선대비분석현시，교정전후광보곡선형상완전불동，임지적반사솔도상적광보곡선형상여USGS표준고적곡선유사；전형지물영상질량적지표평개결과표명，교정후도상적평균제도치기본도대우교정전영상적치；신식적적치정호상반，대기교정적효과명현，진일보증실FLAASH모괴각삼수설치적합이성．
Owing to system and occasional errors, atmospheric correction is the basis of application of remote sensing images. Based on FLAASH model in ENVI4.5, the paper implemented atmospheric correction to one scene image of Hyperion data and evaluated the quality of corrected image. Firstly, the boarders between both land cover types were made more clearly from each other in processed images than that in raw images. The local contrast was enhanced from raw images to corrected images. Secondly, double spectral curves of images demonstrated the obvious difference, while spectral curve for forest was similar to that from standard spectral library of USGS. Lastly, both indices of Average Gradient and Shannon were calculated for four typical objects, including forest, water, transportation and urban land cover. Average Gradient gave a good evidence for the effect of atmospheric correction by larger values for corrected images. Shannon had adverse values and gave an indirect check. All in all, image quality evaluation showed that the atmospheric correction method was a likely solution to eliminate the influence of atmosphere on remote data, while the model and parameters selected were corret.