采用张量子空间的高光谱影像多维滤波算法  

Tensor Subspace Based Multiway Filtering Algorithm for Hyperspectral Images

作　　者：郭贤[1] 黄昕[1] 张乐飞[1] 张良培[1] 

机构地区：[1]武汉大学测绘遥感信息工程国家重点实验室,湖北武汉430079

出　　处：《测绘学报》2013年第2期253-259,267,共8页Acta Geodaetica et Cartographica Sinica

基　　金：国家自然科学基金(41101336;41061130553);国家973计划(2009CB723905);教育部博士点新教师基金(20110141120072);中央高校基本科研业务费(3101016)

摘　　要：提出一种基于张量子空间的多维滤波算法,将其应用于高光谱遥感影像降噪。该方法将高光谱影像数据视为三阶张量,引入张量数据表达,通过张量子空间分解将含噪影像投影到信号子空间,根据影像信号与噪声在子空间中分布的不同滤除噪声并保留原始影像的信号成分。利用该算法作用于多组含噪高光谱数据,对比逐波段二维维纳滤波算法、小波降噪算法等传统数字图像降噪算法的结果,试验证明了这种新型降噪算法的有效性。A novel algorithm for hyperspectral image（HSI） denoising which is based on the tensor subspace is proposed.Considering the HSI as 3 order tensor data,this approach introduces a data representation involving multi-dimensional processing and projects such data into the signal subspace by tensor subspace decomposition.The optimization criterion used in this algorithm is the minimization of mean square error between the estimated signal and the desired signal,then the alternating projection algorithm is adopted to determine the optimal filter in each dimension.Comparative studies with conventional denoising methods such as 2D Wiener filtering and channel-by-channel wavelet thresholding show that our algorithm provides better performance using AVIRIS and PHI datasets.

关 键 词：高光谱影像 降噪 张量子空间 特征值分解 

分 类 号：P237[天文地球—摄影测量与遥感]