FastHyMix: Fast and Parameter-free Hyperspectral Image Mixed Noise Removal

TL;DR

FastHyMix is a rapid, parameter-free method for removing complex mixed noise from hyperspectral images, leveraging Gaussian mixture models, low-rank spectral features, and deep priors for improved denoising performance.

Contribution

It introduces a novel hyperspectral denoising approach combining Gaussian mixture modeling, low-rank spectral analysis, and deep image priors, all without requiring parameter tuning.

Findings

Significant noise reduction in synthetic datasets

Improved image quality on real hyperspectral data

Outperforms existing state-of-the-art denoisers

Abstract

Hyperspectral imaging with high spectral resolution plays an important role in finding objects, identifying materials, or detecting processes. The decrease of the widths of spectral bands leads to a decrease in the signal-to-noise ratio (SNR) of measurements. The decreased SNR reduces the reliability of measured features or information extracted from HSIs. Furthermore, the image degradations linked with various mechanisms also result in different types of noise, such as Gaussian noise, impulse noise, deadlines, and stripes. This paper introduces a fast and parameter-free hyperspectral image mixed noise removal method (termed FastHyMix), which characterizes the complex distribution of mixed noise by using a Gaussian mixture model and exploits two main characteristics of hyperspectral data, namely low-rankness in the spectral domain and high correlation in the spatial domain. The Gaussian mixture model enables us to make a good estimation of Gaussian noise intensity and the location of sparse noise. The proposed method takes advantage of the low-rankness using subspace representation and the spatial correlation of HSIs by adding a powerful deep image prior, which is extracted from a neural denoising network. An exhaustive array of experiments and comparisons with state-of-the-art denoisers were carried out. The experimental results show significant improvement in both synthetic and real datasets. A MATLAB demo of this work will be available at https://github.com/LinaZhuang for the sake of reproducibility.