IEEE Geoscience and Remote Sensing Magazine - June 2013 - 9

not yet operational. The spatial resolutions are higher for
sensors carried by low altitude platforms and vice-versa.
The spectral coverage of HYDICE, AVIRIS, HYPERION,
EnMAP, PRISMA and HyspIRI corresponds to the visible, the near-infrared, and the shortwave infrared spectral bands, whereas CHRIS covers the visible bands and
IASI covers the mid-infrared and the long-infrared bands.
The number of bands is approximately 200 for HYDICE,
AVIRIS, HYPERION, EnMAP, PRISMA and HyspIRI, with
a spectral resolution of the order of 10 nm. The number of
bands for CHRIS is 63, with spectral resolutions of 4 and 12
nm (depending on the region of the spectrum) and 8461
for IASI, with a resolution of 0.5 cm -1. In any case, the resolution is very high (offering a huge potential to discriminate materials) in the case of the first seven sensors, and to
estimate physical parameters (temperature, moisture and
trace gases across the atmospheric column), in the case of
the IASI sensor. A summary of the characteristics of several
hyperspectral imaging instruments currently in operation,
under construction, and missions in a planning stage has
been recently provided [11].
Several factors make the analysis of hyperspectral data
an often complex and hard task calling for sophisticated
methods and algorithms. Among these factors, we refer to
spectral mixing (linear and nonlinear), and degradation
mechanisms associated to the measurement process (e.g.,
noise and atmosphere). Another important issue is the
extremely high dimensionality and size of the data, resulting from the improved spatial, spectral and temporal resolutions provided by hyperspectral instruments. This demands
fast computing solutions that can accelerate the interpretation and efficient exploitation of hyperspectral data sets
in various applications [12]. For example, it has been estimated by the NASA's Jet Propulsion Laboratory (JPL) that
a volume of 4.5 TBytes of data will be daily produced by
HyspIRI (1630 TBytes per year). Similar data volume ratios
are expected for EnMAP and PRISMA. Unfortunately, this
extraordinary amount of information jeopardizes the use
of latest-generation hyperspectral instruments in real-time
or near real-time applications, due to the prohibitive delays
in the delivery of Earth Observation payload data to ground
processing facilities [13]. In this respect, the European Space

DN
Sensor Transfer
Function

Radiometric
Calibration

Sensor
TOA Radiance

Atmospheric
Correction

Atmosphere

Atmospheric RTM

Ground-Leaving Reflectance
BRDF

Surface

Viewing Geometry
and
Surface Correction

Surface Reflectance
FIGURE 3. Spectral characterization of hyperspectral data.

Agency (ESA) already flagged up in 2011 that "data rates and
data volumes produced by payloads continue to increase, while the
available downlink bandwidth to ground stations is comparatively
stable" [14]. In this context, the design of solutions aimed
at taking advantage of the ever increasing dimensionality
of remotely sensed hyperspectral images for near real-time
applications has gained significant relevance and momentum during the last decade [15], [16].
This paper presents a tour over relevant and distinctive
hyperspectral data analysis themes, organized in six main
topics: data fusion, unmixing, classification, target detection, physical parameter retrieval, and fast computing. Most
of the frameworks used in these topics are rooted on signal
and image processing, statistical inference, and machine
learning fields. In all topics, we describe the state-of-the-art
and point to the most likely future challenges and research
directions. Illustrative examples with real data are provided
for some of the topics covered.
The remainder of the paper is organized as follows.
Section II discusses processing techniques aimed at fusing spatial and spectral information from multiple
observation and sources. Section III addresses linear and
nonlinear hyperspectral mixing and unmixing. Section
IV outlines some of the main techniques and challenges

TABLe 1. PARAMeTeRS OF eIGHT HYPeRSPeCTRAL InSTRuMenTS.
PARAMeTeR

HYDICe

AVIRIS

HYPeRIOn

enMAP

PRISMA

CHRIS

HyspIRI

IASI

altitude (km)

1.6

705

653

614

556

626

817

spatial resolution (m)

0.75

5-30

V: 1-2 km
H: 25 km

spectral resolution (nm)

7-14

6.5-10

1.3-12

4-12

0.5 cm -1

coverage (µm)

0.4-2.5

0.4-1.0

0.38-2.5
and 7.5-12

3.62-15.5
(645-2760
cm -1)

number of bands

210

224

220

228

238

217

8461

data cube size
(sample # lines # bands)

200 # 320
# 210

512 # 614
# 224

660 # 256
# 220

1000 # 1000 400 # 880
# 228
# 238

748 # 748
# 63

620 # 512
# 210

765 # 120
# 8461

june 2013

ieee Geoscience and remote sensinG maGazine

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - June 2013