基于差分搜索的高光谱图像解混算法  

Unmixing of hyperspectral images based on differential search

作　　者：张立毅[1,2] 刘静光 陈雷[3,2] 李锵[1] 孙彦慧[1] Zhang Liyi;Liu Jingguang;Chen Lei;Li Qianla;Sun Yanhui(School of Electronic Information Engineering,Tianjin University, Tianjin 300072, China;School of Precision Instrument & Opto-Electronics Engineering, Tianjin University, Tianjin 300072, China;School of Information Engineering, Tianjin University of Commerce, Tianjin 300134 , China)

机构地区：[1]天津大学电子信息工程学院,天津300072 [2]天津商业大学信息工程学院,天津300134 [3]天津大学精密仪器与光电子工程学院,天津300072

出　　处：《计算机应用研究》2016年第10期3177-3180,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(61401307);天津市应用基础与前沿技术研究计划资助项目(15JCYBJC17100);中国博士后科学基金资助项目(2014M561184)

摘　　要：针对高光谱图像解混问题进行研究,发现高光谱图像中各个端元的分布不完全独立,不能将盲源分离方法直接应用于高光谱图像解混。为此,提出了一种基于差分搜索的高光谱图像解混算法。该算法根据高光谱图像丰度非负和丰度和为一特性构造相应的约束项,与互信息相结合作为目标函数,利用差分搜索算法对该目标函数进行优化求解来实现高光谱图像解混。仿真数据和实际数据实验表明,该算法能够有效解决高光谱图像解混问题,与已有其他算法相比,能避免陷入局部极值,提高了图像解混的精度,并且针对不含纯像元的高光谱图像具有很好的解混效果。With regard to the issues of hyperspectral unmixing, the distribution of endmembers were not completely independentin hyperspectral images, thus could not directly apply blind source separation to hyperspectral unmixing. This paper proposeda novel hyperspectral unmixing algorithm based on differential search. According to the abundance non-negative and abundancesum-to-one features, this algorithm constructed corresponding constraint terms and combined it with mutual informationas an objective function, and then optimized the function through differential search algorithm to realize hyperspectral unmixing.The experimental results on simulated and real hyperspectral data demonstrate that the proposed algorithm can effectivelysolve the problem of hyperspectral unmixing. Compared with other algorithms, it can avoid falling into local extremumand get more accurate results, and also be used to unmix hyperspectral data without pure pixels.

关 键 词：高光谱图像解混 差分搜索算法 盲源分离 丰度非负约束 丰度和为一约束 互信息 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]