基于IDL和MATLAB混合编程的两种光谱混合分析方法比较

发布时间：2018-04-02 07:31

本文选题：高光谱影像　切入点：端元变化　出处：《遥感技术与应用》2017年05期

【摘要】：交互式数据语言IDL是一种跨平台、面向对象的应用程序开发语言,具有强大的数据分析和可视化能力。基于矩阵运算的科学计算软件MATLAB是一种大型数值编程计算和图像处理软件,拥有丰富的工具箱。将IDL和MATLAB进行混合编程,同时采用扩展线性混合模型对Hyperion影像进行端元变化解混。为了验证端元变化解混的结果,采用全约束最小二乘法模型进行对比分析。试验结果表明:IDL和MATLAB混合编程结合两者的优势,有效地提高了代码编写效率;同时,当地物类别所占面积较大时,扩展线性混合模型解混精度比全约束最小二乘法模型高,反之,全约束最小二乘法模型的解混精度高。
[Abstract]:Interactive data language (IDL) is a cross-platform, object-oriented application development language with powerful data analysis and visualization capabilities.MATLAB, a scientific computing software based on matrix operation, is a large numerical programming and image processing software with rich toolbox.The IDL and MATLAB are mixed and the extended linear mixed model is used to unmix the Hyperion images.In order to verify the results of endmember unmixing, a fully constrained least square model is used to compare and analyze.The experimental results show that the hybrid programming of MATLAB and MATLAB can effectively improve the coding efficiency, and that the extended linear mixed model has a higher resolution than the fully constrained least square model when the area of local objects is large.On the other hand, the fully constrained least square model has high precision of unmixing.
【作者单位】：中国科学院新疆生态与地理研究所;中国科学院大学;西华师范大学国土资源学院;
【基金】：国家自然科学基金资助项目(41101348) 四川省教育厅自然科学重点项目(15ZA150;17AZ0387) 西华师范大学博士科研启动基金(412546;412547) 西华师范大学英才基金项目(17YC124)
【分类号】：TP751

【相似文献】