Progressive Sample Processing of Band Selection for Hyperspectral Image Transmission

Keng-Hao Liu,Meng-Han Lu,Shih-Yu Chen,Hung-Chang Chien

doi:10.3390/rs10030367

Keng-Hao Liu, Meng-Han Lu + Show 2 more

Open Access

https://doi.org/10.3390/rs10030367

Copy DOI

Abstract

Band selection (BS) is one of the important topics in hyperspectral image (HSI) processing. Many types of BS algorithms were proposed in the last decade. However, most of them were designed for off-line use. They can only be used with pre-collected data, and are sometimes ineffective for applications that require timeliness, such as disaster prevention or target detection. This paper proposes an online BS method that allows us obtain instant BS results in a progressive manner during HSI data transmission, which is carried out under band-interleaved-by-sample/pixel (BIS/BIP) format. Such a revolutionary method is called progressive sample processing of band selection (PSP-BS). In PSP-BS, BS can be done recursively pixel by pixel, so that the instantaneous BS can be achieved without waiting for all the pixels of an image. To develop a PSP-BS algorithm, we proposed PSP-OMPBS, which adopted the recursive version of a self-sparse regression BS method (OMPBS) as a native algorithm. The experiments conducted on two real hyperspectral images demonstrate that PSP-OMPBS can progressively output the BS with extremely low computing time. In addition, the convergence of BS results during transmission can be further accelerated by using a pre-defined pixel transmission sequence. Such a significant advantage not only allows BS to be done in a real-time manner for the future satellite data downlink, but also determines the BS results in advance, without waiting to receive every pixel of an image.

Highlights

Due to the use of hundreds of spectral bands, hyperspectral imaging (HSI) generally has enormous data volume and contains vast amount of information
This paper presents an instant band selection (BS) method based on progressive sample processing (PSP), called PSP-orthogonal matching pursuit-based BS (OMPBS)
Unlike traditional BS methods, which must re-implement the total received data, PSP-OMPBS can immediately obtain BS results when receiving a new pixel by referring to past information

Summary

Introduction

Due to the use of hundreds of spectral bands, hyperspectral imaging (HSI) generally has enormous data volume and contains vast amount of information. This special characteristic results in several issues. The inter-band correlation of HSI is very high, and adjacent bands may contain redundant spectral information. This would lead to the well-known problem called the “curse of dimensionality” in data analysis. Under such circumstances, removing partial data without significant loss of an image’s spectral information is necessary. BS takes advantage of such high-band correlation to remove the redundant bands, in order to achieve a wide range of applications, such as dimensionality reduction, data storage, data transmission, target detection, and classification

Methods

Findings

Discussion

Conclusion