中文摘要：

水稻中过量砷（As）能够损害叶片中叶绿素和叶片内部结构,进而影响水稻光合作用效率,并改变水稻在光谱上的表现。利用高光谱植被指数（CARI、PRI、SIPI）和独立变量分析（ICA）模型对水稻中As含量进行了研究。结果表明,以上3种高光谱植被指数与水稻中As含量均呈一定的相关关系,其相关系数在0.67以上;而经过独立变量分析（ICA）可知,在蓝光波段（440～540nm）和红光波段（600～700nm）之间各有一个独立变量与水稻中As含量高度相关,相关系数达到0.95以上。将上述植被指数与独立变量和水稻中As含量之间进行回归分析,得到水稻中As含量的线性回归方程。研究表明,重金属As对水稻生长的影响可以通过其在光谱上的特征（如相关植被指数）改变来体现,并可以用独立变量分析（ICA）方法提取光谱中关于As胁迫的隐含弱信息,建立遥感预测模型,为大面积监测农作物As污染提供依据。

英文摘要：

High arsenic content in rice can damage the chlorophyll and structure in their leaves,reduce the rate of photosynthesis and change their spectral features.The model established in this paper by the independent component analysis（ICA）and hyperspectral vegetation indices（CARI,PRI,SIPI）were used to predict the arsenic content in mature rice.Sixty samples belonging to mature rice in three different areas were scanned by ASD field pro3.Arsenic reference values were obtained by atomic absorption spectrometry.The study showed that there was linear correlation between these three hyperspectral vegetation indices（CARI,PRI,SIPI）calculated from ASD optical data and the content of arsenic in mature rice,whose coefficients（R）are-0.67,0.82,0.91,using least-squares regression.The fast independent component analysis（ICA）was applied in a matrix 20×60,which represents the 20 groups optical data of the 60 samples.The 20 groups optical data previously centralized and standardized were respectively divided from the 100 nm range of blue band（440～540 nm）and red band（600～700 nm）by 5 nm intervals,and that produced four variables（IC-B1,IC-B2;IC-R1,IC-R2）from red band and blue band.Furthermore,IC-B1 and IC-R1 had high correlation with the arsenic content in mature rice（R=0.96,R=-0.95）.So the model established in this paper by those variables and hyperspectral vegetation indices can predict arsenic contamination in mature rice.