IEEE Geoscience and Remote Sensing Magazine - December 2017 - 41

HSIs and, at the same time, overcome the challenges caused
by the hyperspectral data set. We elaborate on different CD
methods in the "Change Detection" section.
Another crucial aspect of HSI analysis to be precisely
taken into account is that hyperspectral remote sensors
are now in the era of massive automatic data collection
resulting from the improved spatial, spectral, and temporal resolutions provided by several hyperspectral instruments. As a result, fast computing (detailed in the "Fast
Computing" section) is critical to accelerating the efficient
exploitation and analysis of HSIs.
MISSIONS AND STATISTICS
Several hyperspectral imaging instruments are currently
available for the purpose of remote-sensing image and
signal analysis, providing a large volume of images for
various thematic applications. Airborne hyperspectral
imaging sensors [e.g., the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS), Hyperspectral Digital Imagery Collection Experiment, Compact Airborne Spectrographic Imager (CASI), Airborne Prism Experiment, and
HySpex] play a central role in acquiring data sources for
the hyperspectral scientific community. The European
Facility for Airborne Research (http://www.eufar.net) has
established standards and protocols in the field of airborne hyperspectral remote sensing, allowing transnational access to national infrastructures. In recent years,
unmanned-aerial-vehicle-based real-time hyperspectral
imaging has become increasingly common in various applications, such as agricultural monitoring, while raising
new challenges in image processing.
Table 1 presents the principal parameters of seven spaceborne imaging spectroscopy missions planned for the near
future: the China commercial remote-sensing satellite system, DLR Earth sensing imaging spectrometer (DESIS) [5],
environmental mapping and analysis program [6], hyperspectral imager suite (HISUI) [7], precursore iperspettrale
della missione applicativa (PRISMA) [8], spaceborne hyperspectral applicative land and ocean mission (Shalom)
[9], and hyperspectral infrared imager [10]. Many of those
satellites are designed to have a GSD of 30 m, aiming at

global coverage. Shalom's GSDs of 8 m and 10 m in the
visible near-infrared (VNIR) and SWIR ranges, respectively,
are driven by both operational and commercial needs. DESIS and HISUI will be mounted on the International Space
Station. The launch of these satellites will further accelerate
research on HSI processing and its applications.
Figure 3 shows statistics on articles related to HSIs and
signal processing published in IEEE journals during 2009-
2012 and 2013-2016. All articles were searched via IEEE
Xplore using hyperspectral as
the main keyword and then
categorized into broad topics
UNMANNED-AERIALby analyzing keywords in the
VEHICLE-BASED REAL-TIME
titles. The size of each pie in
HYPERSPECTRAL IMAGING
Figure 3 is proportional to the
HAS BECOME INCREASINGLY
number of articles.
COMMON IN VARIOUS
The totals returned by this
APPLICATIONS.
search are an indicator of the
hyperspectral community's
recent growth. Seven topics
under investigation represented 61.5% of all of the articles
published in 2013-2016. Classification was the most actively addressed topic in both periods, while spectral unmixing
was the second most common. Classification- and unmixing-related studies accounted for 41.6% of the total. These
top two topics were followed by DR and image restoration.
Image restoration showed a high growth rate, indicating
that the improvement in the quality of HSIs is significant
in subsequent processing. Resolution enhancement received
particular attention during 2013-2016, as demonstrated by
the highest growth rate. Although the number of articles related to CD increased steadily, the overall number was still
small, probably due to limited data sets.
CONTRIBUTION
This article introduces a detailed and organized overview of
HSIs and signal processing, categorized into the seven different themes previously mentioned. In each section, we provide some numerical results, illustrations, a critical overview
of the state of the art, current challenges, and possible future
works. It is worth noting that the methodologies described

TABLE 1. THE PARAMETERS OF SEVEN SPACEBORNE IMAGING SPECTROSCOPY MISSIONS.
PARAMETER

CCRSS

DESIS

EnMAP

HISUI

PRISMA

Shalom

HyspIRI

Altitude (km)

400

653

400

615

600

626

GSD (m)

Bandwidth (nm)

5-20

3.3

5.25-12.5

10-12.5

# 12

# 10

Spectral coverage ( μm)

0.4-2.5

0.4-1.0

0.42-2.45

0.44-2.5

0.4-2.5

0.38-2.5

Number of bands

328

180

228

185

237

241

210

Swath width (km)

30.7

30-60

Other sensor

Pan

TIR

TIR: thermal infrared; CCRSS: China commercial remote-sensing satellite system; EnMAP: environmental mapping and analysis program; HyspIRI: -hyperspectral infrared imager.

DECEMBER 2017

IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE

http://www.eufar.net

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - December 2017