Hyperspectral Image Denoising Using First Order Spectral Roughness Penalty in Wavelet Domain

被引：66

作者：

Rasti, Behnood ^{[1
]}

Sveinsson, Johannes R. ^{[1
]}

Ulfarsson, Magnus Orn ^{[1
]}

Benediktsson, Jon Atli ^{[1
]}

机构：

[1] Univ Iceland, Dept Elect & Comp Engn, IS-107 Reykjavik, Iceland

来源：

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING | 2014年 / 7卷 / 06期

关键词：

Classification; denoising; first order roughness penalty; hyperspectral image; multiresolution analysis; penalized least squares; Stein's unbiased risk estimator; wavelets; SPATIAL CLASSIFICATION; ATTRIBUTE PROFILES;

D O I：

10.1109/JSTARS.2013.2272879

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

In this paper, a new denoising method for hyperspectral images is proposed using First Order Roughness Penalty (FORP). FORP is applied in the wavelet domain to exploit the Multi-Resolution Analysis (MRA) property of wavelets. Stein's Unbiased Risk Estimator (SURE) is used to choose the tuning parameters automatically. The simulation results show that the penalized least squares using FORP can improve the Signal to Noise Ratio (SNR) compared to other denoising methods. The proposed method is also applied to a corrupted hyperspectral data set and it is shown that certain classification indices improve significantly.

引用

页码：2458 / 2467

页数：10

共 33 条

[1]

[Anonymous], WAVELET TOUR SIGNAL

[2]

[Anonymous], MULTIVARIATE IMAGE P

[3]

[Anonymous], 2003, P VLSM

[4]

[Anonymous], 2006, P SPIE