基于端元变化的两种混合像元分解算法比较研究  被引量:5

Comparison Analysis Between Two Spectral Mixture Analysis Methods of Incorporating Endmember Variability

在线阅读下载全文

作  者:段金亮[1] 王杰[1] 文星跃[1] 

机构地区:[1]西华师范大学国土资源学院,四川南充637009

出  处:《资源开发与市场》2017年第6期651-655,F0002,共6页Resource Development & Market

基  金:国家自然科学基金面上项目(编号:41671220);西华师范大学博士科研启动基金项目(编号:412546;412547);四川省教育厅自然科学基金重点项目(编号:15ZA150)

摘  要:光谱混合分析对提高遥感影像分类具有重要意义,其中端元变化处理是提高解混精度的关键。目前,许多算法被用来解决端元变化,但仍存在一些问题有待解决,如算法运行效率慢、忽略端元的高阶交互、像元空间邻城信息缺失。结合IDL和MATLAB混合编程,利用确定性模型中的交替最小角度法和统计性模型中考虑高阶项的非线性算法对Hyperion影像进行端元变化解混,再利用概率松弛标记法对像元补充空间邻域信息。试验结果表明:当某种地物类别所占面积较大时,确定性与统计性模型都能获得较高的解混精度;当地物类别所占面积较小时,确定性模型的解混精度高于统计性模型;补充像元空间邻域信息对解混结果有很好的校正。Spectral mixing analysis was of great significance to improve the classification of remote sensing images. The processing of the change in endmember was the key to improving the solution. Many algorithms were proposed to solve the problem, but there were still some problems. For ex- ample, the poor operation efficiency of the algorithm, ignoring the high orer interaction of the endmembers, and missing some information in the spatial domain. To solve problems above, based on the combination of IDL and MATLAB mixed programming, this paper used the alternate deterministic mod- el minimum angle method and statistical model for the sake of the nonlinear algorithm of high order terms of endmember unmixing changes within the Hyperion images, and the probabilistic relaxation labeling on pixel space field information was used to improve the unmixing accuracy. The experimen- tal results showed that both the deterministic model and the statistical model could obtain a high degree of unmixing accuracy when a certain object category occupied a large area in a image. When the area occupied by the object category was small, the deterministic model's accuracy was superior to the statistical model. In addition, when the field information of the pixel space was supplemented, the result of the unmixing could be corrected.

关 键 词:HYPERION影像 端元变化解混 IDL与MATLAB混合编程 交替最小角度法 非线性混合模型 

分 类 号:TP751.1[自动化与计算机技术—检测技术与自动化装置]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象