摘要
从理论上阐述了混合像元的产生机理和混合像元线性分解模型,研制了一套地面成像光谱实验系统,直接获取了高精度的端元光谱数据,通过像元合并的方式构造混合像元,针对不同丰度的非立体空间混合像元端元,分析了根据线性分解模型加权计算得到的光谱与实际混合光谱的误差,进而论证了线性分解模型的精度.
The theoretical analysis of emergence mechanism and linear unmixing model of mixed pixel was presented.A novel compact field hyperspectral imaging system was setup,and high accurate spectrum of endmembers was collected directly.Based on these endmembers,the model of linear unmixing of the mixed pixels was studied.The weighted spectrum obtained with the model for endmember with different abundance in the whole non-three-dimensional space was compared with the practical mixing spectrum.The accuracy of the model was verified.
出处
《红外与毫米波学报》
SCIE
EI
CAS
CSCD
北大核心
2010年第5期357-361,共5页
Journal of Infrared and Millimeter Waves
基金
国家重点基础研究发展规划(973)项目(2009CB724005)
关键词
超光谱成像
混合像元分解
线性分解模型
端元
hyperspectral imaging
mixed pixel unmixing
linear unmixing model
endmember
