CAJ | 학술논문

利用研究区地物光谱曲线的光谱吸收指数、光谱斜率，光谱吸收深度等光谱特征参量建立多变量决策树，成功识别出白云母、方解石、绿泥石、绿帘石、蛇纹石等多种矿物。精度分析表明：建立的决策树识别结果与遥感中心的填图结果基本一致，尤其是绿帘石和蛇纹石填图结果吻合的较好，而白云母、绿泥石、方解石矿物填图结果稍有差异。
이용연구구지물광보곡선적광보흡수지수、광보사솔，광보흡수심도등광보특정삼량건립다변량결책수，성공식별출백운모、방해석、록니석、록렴석、사문석등다충광물。정도분석표명：건립적결책수식별결과여요감중심적전도결과기본일치，우기시록렴석화사문석전도결과문합적교호，이백운모、록니석、방해석광물전도결과초유차이。
This paper establishes a multi-variable decision tree by using spectral absorption indexes, spectral slopes, spectral absorption depths and other spectral characteristics, and successfully identifies various minerals, including muscovite, calcite, chlorite, epidote, and serpentine. The precision analysis shows that the identification results of the decision tree established in this paper are consistent with the mapping results of China Aero Geophysical Survey ＆ Remote Sensing Center for Land and Resources, especially the mapping results of epidote and serpentine, with slight difference in the mapping results of muscovite, chlorite, and calcite.