摘要
针对传统线性光谱分解中端元选择范围大、信息损失多的问题,对端元选择做了改进.以上海市中心区域为例,进行了改进研究,并采用高分辨率航空遥感影像进行了检验.结果显示,引入像元纯净指数,为端元的选择提供了基本的定量准则;同时构造了三维特征空间动态交互模式,则从技术上改变了端元选择的模式.均方根以及高分辨率航空影像的抽样检验都表明,混合像元分解的精度较好,说明该改进方法对于城市混合像元分解是可行的.
Aimed at over big bound and overmuch information loss of endmember selection, traditional unmixing techniques were improved from the aspects of endmember selection. Improved linear unmixing technique was applied to the center urban area of Shanghai and tested by high resolution aerial photograph. The unmixing process showed that a basic quantificational rule was provided for endmember selection by introducing pixel purity index and that the mode of endmember selection was changed technically by build ing a 3D feature space. Root mean square （RMS） error and sampling inspection of high-resolution aerial photograph show that pixel unmixing has higher precision and that the improved technique is appropriate for urban pixel unmixing.
出处
《浙江大学学报（工学版）》
EI
CAS
CSCD
北大核心
2006年第4期719-723,共5页
Journal of Zhejiang University：Engineering Science
基金
国家自然科学基金资助项目(10371092)
上海市城市化生态过程与生态恢复重点实验室开放基金资助项目
关键词
像元分解
端元选择
盖度
上海
pixel unmixing
endmember selection
fraction
Shanghai
作者简介
岳文泽（1977-）,男,安徽凤台人,博士后,从事遥感与城市土地利用研究. E-mail： wzyue@zju.edu.cn
