Superpixel Spectral Unmixing for Hyperspectral Image Superresolution Using a Coupled Encoder-Decoder Network

Shao-Lei Zhang,Yu-Qing Zhao,Hong-Qiao Wang,Guang-Yuan Fu,Carlos Michel

doi:10.1155/2020/8886178

Shao-Lei Zhang, Yu-Qing Zhao + Show 3 more

Open Access

https://doi.org/10.1155/2020/8886178

Copy DOI

Abstract

In this paper, we propose a novel hyperspectral image superresolution method based on superpixel spectral unmixing using a coupled encoder-decoder network. The hyperspectral image and multispectral images are fused to generate high-resolution hyperspectral images through the spectral unmixing framework with low-rank constraint. Specifically, the endmember and abundance information is extracted via a coupled encoder-decoder network integrating the priori for unmixing. The coupled network consists of two encoders and one shared decoder, where spectral information is preserved through the encoder. The multispectral image is clustered into superpixels to explore self-similarity, and then, the superpixels are unmixed to obtain an abundance matrix. By imposing a low-rank constraint on the abundance matrix, we further improve the superresolution performance. Experiments on the CAVE and Harvard datasets indicate that our superresolution method outperforms the other compared methods in terms of quantitative evaluation and visual quality.

Highlights

With rich spectral and spatial information, hyperspectral images (HSI) have received extensive attention and have been widely used in many fields, especially for remote sensing [1, 2] and medical imaging [3]
There were some HSI superresolution methods based on sparse representation theory that is widely used in computer vision tasks
In order to evaluate the accuracy of the estimated highresolution HSI, four widely used quality metrics are used, including root mean square error (RMSE), peak signal-tonoise ratio (PSNR), spectral angle mapper (SAM), and relative dimensionless global error in synthesis (ERGAS)

Summary

Introduction

With rich spectral and spatial information, hyperspectral images (HSI) have received extensive attention and have been widely used in many fields, especially for remote sensing [1, 2] and medical imaging [3]. Due to hardware limitations, hyperspectral imaging requires long exposure times to ensure a high signal-to-noise ratio, which leads to low spatial resolution [4]. Image superresolution is a promising way to acquire high-resolution images in both spatial and spectral domains. There were some HSI superresolution methods based on sparse representation theory that is widely used in computer vision tasks. Akhtar et al [4] impose a nonnegative constraint on spectral dictionary and sparse representations to improve HSI superresolution performance. Han et al.’s work [7] combined nonnegative sparse representation, local similarity, and nonlocal similarity for HSI superresolution, which explores self-similarity in superpixel and across the entire image. When learning the spectral dictionary, these methods do not take into account the spectral correlation in the HSI causing severe spectral distortion in the reconstructed images

Methods

Results

Conclusion