Noise Reduction in Hyperspectral Imagery: Overview and Application

Abstract

Hyperspectral remote sensing is based on measuring the scattered and reflected electromagnetic signals from the Earth's surface emitted by the Sun. The received radiance at the sensor is usually degraded by atmospheric effects and instrumental (sensor) noises which include thermal (Johnson) noise, quantization noise, and shot (photon) noise. Noise reduction is often considered as a preprocessing step for hyperspectral imagery. In the past decade, hyperspectral noise reduction techniques have evolved substantially from two dimensional bandwise techniques to three dimensional ones, and varieties of low-rank methods have been forwarded to improve the signal to noise ratio of the observed data. Despite all the developments and advances, there is a lack of a comprehensive overview of these techniques and their impact on hyperspectral imagery applications. In this paper, we address the following two main issues; (1) Providing an overview of the techniques developed in the past decade for hyperspectral image noise reduction; (2) Discussing the performance of these techniques by applying them as a preprocessing step to improve a hyperspectral image analysis task, i.e., classification. Additionally, this paper discusses about the hyperspectral image modeling and denoising challenges. Furthermore, different noise types that exist in hyperspectral images have been described. The denoising experiments have confirmed the advantages of the use of low-rank denoising techniques compared to the other denoising techniques in terms of signal to noise ratio and spectral angle distance. In the classification experiments, classification accuracies have improved when denoising techniques have been applied as a preprocessing step.

Influence of the System MTF on the On-Board Lossless Compression of Hyperspectral Raw Data

“…Let us notice that the radiometric range of the Aviris 2006 raw data is of 16 bits, so that the range of the RMSE values is proportionally high. 3 3.5 Figure 13 presents several plots, which are similar to the one of Figure 12, but with different standard deviations of the noise added to the filtered images before compression. Evidently, the addition of more signal-independent noise contributions does not substantially change the trend of the plot in Figure 12.…”

ADRN: Attention-Based Deep Residual Network for Hyperspectral Image Denoising

“…Hyperspectral image (HSI) data contains abundant saptial and spectral information, which makes it have a wide range of applications. Nevertheless, because of the senosr restriction and atmospheric interference, HSIs often suffer from various types of noise, such as Gaussian noise, stripe noise and dead lines, etc [1]. Thus, it is essential to reduce the noise in HSIs in order to facilitate the following high-level analysis tasks.…”

Hyperspectral Image Denoising using Dictionary Learning

“…In particular, for remotely acquired HSI, at least two kinds of noise are of importance: noise due to the spectral sensor sensitivity, and noise due to the swiping pattern of the sensors which yields stripes. Moreover, in the presence of clouds or other atmospheric perturbation, missing data may be present as well [2]. In this work, we will suppose that HSI are only corrupted with anisotropic Gaussian noise in order to simplify the analysis of our denoising method, but other types of noise as well as missing data can be tackled with similar tools.…”

“…Removing noise in HSI is an important pre-processing step for any learning task such as segmentation, detection or spectral unmixing, and has therefore been studied extensively in the literature. State-of-the-art techniques exploit two key properties [2]: HSI typically are low rank matrices once vectorized in the spatial dimensions; and HSI (or small patches of it) are sparse in some well chosen bases such as wavelets. In particular, the state-of-the-art method HyRes [3] makes use solely of these two assumptions to very efficiently denoise remotely acquired HSI with minimal computation time.…”