基于高光谱分形分析的水稻镉污染动态监测

Dynamically Monitoring Rice Cd Pollution Based on Hyperspectral Fractal Analysis

水稻中过量镉会影响叶绿素含量和细胞结构,进而改变水稻在光谱维上的特征表现。利用长春郊区实测大田水稻高光谱数据和同步测量的土壤、水稻生化参数,系统分析水稻冠层光谱曲线的分形特征及不同镉污染胁迫程度下光谱分维与两类光谱指数的变化情况,构建光谱分维、光谱指数与水稻叶片镉浓度的空间分布图,探讨光谱分维与光谱指数在提取镉污染胁迫等级信息中的相互关系及物理机制。研究表明,光谱分维相对于光谱指数能够更好地综合反映镉污染胁迫下水稻生理特征参数的变化,因而为有效地大范围动态监测水稻镉污染提供诊断依据。

Excess of Cd in rice will affect the chlorophyll content and cell structure,and then change the reflectance spectra.In this paper,hyperspectral data,collected by the ASD portable spectrometer in rice field in suburb of Changchun,and the soil-rice characteristic parameters measured simultaneously are used to analysis the fractal characteristic of rice canopy spectral curve and how fractal dimension and two kinds of spectral indices change under different levels of Cd pollution.Then the spatial distribution maps of fractal dimension,hyperspectral indices and Cd density in rice leaves were built,and relationship and physical mechanisms of fractal dimension and two kinds of spectral indices were analyzed.Studies have shown that fractal dimension of spectral curve can reflect the change of physical characteristic parameters integrally better than spectral indices.Therefore it can provide diagnosis basis for effective large-scale monitoring of rice Cd contamination.