Method of Multispectral Image Denoising Based on Whole and Sub-Sparsity

Wenjia Zeng,Xinggan Zhang

doi:10.1109/access.2021.3076786

Wenjia Zeng, Xinggan Zhang

Open Access

https://doi.org/10.1109/access.2021.3076786

Copy DOI

Journal: IEEE Access	Publication Date: Jan 1, 2021
Citations: 32	License type: CC BY 4.0

Affiliation: Nanjing University

Abstract

Multi-Spectral Image(MSI) denoising is an important preprocessing procedure to improve the performance of high-level processing. Tensor-based approach is one of the most popular methods for MSI denoising, since MSIs can be seen as multi-dimension arrays containing both spatial and spectral information. There are two main information in MSI, Global Correlation along Spectrum(GCS) and Nonlocal Self Similarity across space(NSS). Most tensor based approaches exploited these two characteristics by low-rank regularizations, mainly based on CANDERCOMP/PARAFAC(CP) decomposition and Tucker decomposition. However, they did not show a clear physical meaning. In this paper, we exploit the fact that pixels in MSI often cover several different materials and so that tensor data is mixed. Based on this, we divide tensor into several sub-tensors and propose a novel low rank regularization called Whole and Sub-Sparsity(WSS): GCS is modeled in the sub-tensors and NSS is modeled in the original tensor, which shows a clear physical meaning. Besides, to solve our model, we develop the corresponding algorithm by employing alternating direction method of multipliers(ADMM) framework. Experiment results show that our method is competitive compared to all state of the art MSI denoising methods.

Highlights

Multi-spectral image(MSI), which contains various images from hundreds of spectral bands, has a wide range of applications in fields such as environmental monitoring [12], military surveillance [33], and agriculture [28]
Tensor data is divided into several sub tensors, and NSS and GCS are measured in the whole tensor data and sub-tensors, respectively; 3) We propose the corresponding algorithm for solving this model based on alternating direction method of multipliers (ADMM) framework
We find non-local neighboring areas for each FBP and reformulate a 3D tensor set: {Yk : Yk ∈ Rdhdw×ds×dn }, where dn is the number of non-local similar FBPs of Yk, and each Yk is the input of Y in algorithm 1

Summary

INTRODUCTION

Multi-spectral image(MSI), which contains various images from hundreds of spectral bands, has a wide range of applications in fields such as environmental monitoring [12], military surveillance [33], and agriculture [28]. The GCS indicates that images over different bands are highly correlated, while the NSS refers to the fact that correlation exists in different local full band patches(FBP) Both spectral correlation and spatial correlation are the intrinsic features. Albeit demonstrated to be effective, most of these methods either did not fully consider both GCS and NSS, or did not provide a clear physical meaning. Zhang: Method of MSI Denoising Based on Whole and Sub-Sparsity modes: spatial mode, spectral mode, non local similar across space mode(see III.B). Tensor data is divided into several sub tensors, and NSS and GCS are measured in the whole tensor data and sub-tensors, respectively; 3) We propose the corresponding algorithm for solving this model based on ADMM framework

RELATED WORK

TENSOR DATA REORGANIZATION

ADMM ALGORITHM

SIMULATED EXPERIMENT

CONCLUSION