Hyperspectral Imaging - MDPI

0 downloads 213 Views 2MB Size Report
Oct 30, 2017 - acquisition devices are addressed, including sensor types, acquisition ..... implies the use of a confine
remote sensing Technical Note

Hyperspectral Imaging: A Review on UAV-Based Sensors, Data Processing and Applications for Agriculture and Forestry Telmo Adão 1,2, * ID , Jonáš Hruška 2 , Luís Pádua 2 Raul Morais 1,2 ID and Joaquim João Sousa 1,2 1 2

*

ID

, José Bessa 2 , Emanuel Peres 1,2

ID

,

Institute for Systems and Computer Engineering, Technology and Science (INESC-TEC—Formerly INESC Porto), 4200-465 Porto, Portugal; [email protected] (E.P.); [email protected] (R.M.); [email protected] (J.J.S.) Department of Engineering, School of Sciences and Technology, University of Trás-os-Montes e Alto Douro, 5000-801 Vila Real, Portugal; [email protected] (J.H.); [email protected] (L.P.); [email protected] (J.B.) Correspondence: [email protected]; Tel.: +351-259-350-356

Received: 20 September 2017; Accepted: 27 October 2017; Published: 30 October 2017

Abstract: Traditional imagery—provided, for example, by RGB and/or NIR sensors—has proven to be useful in many agroforestry applications. However, it lacks the spectral range and precision to profile materials and organisms that only hyperspectral sensors can provide. This kind of high-resolution spectroscopy was firstly used in satellites and later in manned aircraft, which are significantly expensive platforms and extremely restrictive due to availability limitations and/or complex logistics. More recently, UAS have emerged as a very popular and cost-effective remote sensing technology, composed of aerial platforms capable of carrying small-sized and lightweight sensors. Meanwhile, hyperspectral technology developments have been consistently resulting in smaller and lighter sensors that can currently be integrated in UAS for either scientific or commercial purposes. The hyperspectral sensors’ ability for measuring hundreds of bands raises complexity when considering the sheer quantity of acquired data, whose usefulness depends on both calibration and corrective tasks occurring in pre- and post-flight stages. Further steps regarding hyperspectral data processing must be performed towards the retrieval of relevant information, which provides the true benefits for assertive interventions in agricultural crops and forested areas. Considering the aforementioned topics and the goal of providing a global view focused on hyperspectral-based remote sensing supported by UAV platforms, a survey including hyperspectral sensors, inherent data processing and applications focusing both on agriculture and forestry—wherein the combination of UAV and hyperspectral sensors plays a center role—is presented in this paper. Firstly, the advantages of hyperspectral data over RGB imagery and multispectral data are highlighted. Then, hyperspectral acquisition devices are addressed, including sensor types, acquisition modes and UAV-compatible sensors that can be used for both research and commercial purposes. Pre-flight operations and post-flight pre-processing are pointed out as necessary to ensure the usefulness of hyperspectral data for further processing towards the retrieval of conclusive information. With the goal of simplifying hyperspectral data processing—by isolating the common user from the processes’ mathematical complexity—several available toolboxes that allow a direct access to level-one hyperspectral data are presented. Moreover, research works focusing the symbiosis between UAV-hyperspectral for agriculture and forestry applications are reviewed, just before the paper’s conclusions. Keywords: hyperspectral; UAS; UAV; hyperspectral sensors; hyperspectral data processing; agriculture; forestry; agroforestry

Remote Sens. 2017, 9, 1110; doi:10.3390/rs9111110

www.mdpi.com/journal/remotesensing

Remote Sens. 2017, 9, 1110

2 of 30

Remote Sens. 2017, 9, 1110

2 of 31

1. 1. Introduction Introduction Remote Remote sensing sensing relying relying on on unmanned unmanned aircraft aircraft systems systems (UAS), (UAS), although although an an emerging emerging field field of of application, has been systematically applied for monitoring vegetation and environmental parameters application, has been systematically applied for monitoring vegetation and environmental aiming the optimization of agroforestry activities [1].activities In this context, UAS have become suitable to parameters aiming the optimization of agroforestry [1]. In this context, UAS have become assess crops’ conditions by gathering huge amounts of raw data that require further processing suitable to assess crops’ conditions by gathering huge amounts of raw data that require further to enable a to wide range of applications, such as water [2], status vigor [2], assessment [3], biomass processing enable a wide range of applications, suchstatus as water vigor assessment [3], estimation [4] and disease monitoring [5]. Similarly, forestry and nature preservation can also greatly biomass estimation [4] and disease monitoring [5]. Similarly, forestry and nature preservation can benefit frombenefit the usefrom of UAS technology, allowing inspection of forestryofoperations [6], wildfire also greatly the use of UAS technology, allowing inspection forestry operations [6], detection [7], health monitoring [8] and forest preservation [9]. wildfire detection [7], health monitoring [8] and forest preservation [9]. In number of successful works that have been applied to agroforestry and related In spite spiteofofthe thelarge large number of successful works that have been applied to agroforestry and areas using low-cost passive imagery sensors—such as visible (RGB) and near infrared (NIR)—many related areas using low-cost passive imagery sensors—such as visible (RGB) and near infrared applications require higher spectral fidelityspectral that only multispectral and hyperspectral sensors (NIR)—many applications require higher fidelity that only multispectral and[10–13] hyperspectral can offer. Both of the referred spectral-based methods consist of the acquisition of images where, [10–13] sensors can offer. Both of the referred spectral-based methods consist of the acquisition of for each of the image’s spatially distributed elements, a spectrum of the energy reaching the respective images where, for each of the image’s spatially distributed elements, a spectrum of the energy sensor is the measured. The mainis difference is thebetween numberthem of bands referred as reaching respective sensor measured. between The mainthem difference is the (also number of bands channels) and their width [14]. While multispectral imagery generally ranges from 5 to 12 bands that (also referred as channels) and their width [14]. While multispectral imagery generally ranges from are in pixels and each band acquired remote sensing radiometer, hyperspectral 5 torepresented 12 bands that are represented in ispixels andusing each aband is acquired using a remote sensing imagery consists of a much higher band number—hundreds or thousands of them—arranged in radiometer, hyperspectral imagery consists of a much higher band number—hundreds or thousands aofnarrower bandwidth (5–20 nm, each). Figure 1 represents the differences between multi and them—arranged in a narrower bandwidth (5–20 nm, each). Figure 1 represents the differences hyperspectral which rely on a spectroscopic approach that hasapproach been used forhas laboratorial between multiimaging, and hyperspectral imaging, which rely on a spectroscopic that been used practices and in Astronomy for more than 100 years. for laboratorial practices and in Astronomy for more than 100 years.

Figure 1. Spectrum representation including: (A) Multispectral example, with 5 wide bands; and (B) Figure 1. Spectrum representation including: (A) Multispectral example, with 5 wide bands; and Hyperspectral example consisting of several narrow bands that, usually, extends to hundreds or (B) Hyperspectral example consisting of several narrow bands that, usually, extends to hundreds or thousands of them (image not drawn to scale, based in [14]). thousands of them (image not drawn to scale, based in [14]).

In fact, both multi- and hyperspectral imagery have the potential to take data mining to a whole fact, both level multi-inand hyperspectral imagery the assessment potential to [11] takeand dataagriculture mining to [12]. a whole new In exploration many areas, including foodhave quality For new exploration level in many areas, including food quality assessment [11] and agriculture [12]. instance, productivity and stress indicators in both agricultural and forest ecosystems can be assessed For instance, productivity indicators in both which agricultural forestby ecosystems through photosynthetic lightand usestress efficiency quantification, can beand obtained measuringcan the be assessed through photosynthetic use on efficiency quantification, which can be obtained by photochemical reflectance index (PRI)light relying narrowband absorbance of xanthophyll pigments measuring the photochemical reflectance index (PRI) relying on narrowband absorbance of xanthophyll at 531 and 570 nm [15]. However, while the higher spectral resolution present in hyperspectral data pigments at 531 sensing and 570 nm [15]. However, while the higher spectral resolution present in hyperspectral allows remote of narrowband spectral composition—also known as spectra, signature or, data allowstoremote sensing signature—multispectral of narrowband spectral composition—also known as spectra, signature or, according [16], spectral data manifests itself in larger intervals over the according to [16], spectral signature—multispectral data manifests itself in larger intervals over the electromagnetic spectrum, which does not enable to reach the same level of detail. Thus, electromagnetic spectrum, whichperformance does not enable to reach the sameand level of detail. endmembers Thus, hyperspectral hyperspectral data has a better profiling materials respective due to data has a better performance profiling materials and respective endmembers due to its almost its almost continuous spectra. On the one hand, it covers spectral detail that might pass unnoticeable continuous spectra. On the one hand,and it covers detail that might pass 2,unnoticeable in in multispectral data due to its discrete sparsespectral nature. For example, in Figure since red-edge multispectral its accessible discrete andthrough sparse nature. For example, in Figure since red-edge (RE, (RE, 670–780data nm)due is to not the broadband sensor, leaf 2, chlorophyll content, phenological state and vegetation stress—which are parameters that manifest in that spectral range—

Remote Sens. 2017, 9, 1110

3 of 30

670–780 nm) is not accessible through the broadband sensor, leaf chlorophyll content, phenological Remote Sens. 2017, 9, 1110 3 of 31 state and vegetation stress—which are parameters that manifest in that spectral range—cannot be assessed. On the other hyperspectral has the ability to discriminate components that may that be cannot be assessed. Onhand, the other hand, hyperspectral has the ability to discriminate components unwittingly grouped by multispectral bands (see, for example, [17] for more details). may be unwittingly grouped by multispectral bands (see, for example, [17] for more details).

Figure 2. Broadband sensors’ inability for accessing the spectral shift of the RE (670–780 nm) slope Figure 2. Broadband sensors’ inability for accessing the spectral shift of the RE (670–780 nm) slope associated with leaf chlorophyll content, phenological state and vegetation stress, comparatively to associated with leaf chlorophyll content, phenological state and vegetation stress, comparatively to wide-range narrowband ones (re-edited from [18]). wide-range narrowband ones (re-edited from [18]).

Along with the resolution improvement, the hyperspectral sensing approach also increases data Along with the resolution improvement, the hyperspectral sensing approachof also increases data processing complexity, since such imagery ranges from hundreds to thousands narrow bands that processing complexity, since such imagery ranges from hundreds to thousands of narrow bands can be difficult to handle in real-time with reduced computational resources. Besides, spectral that can be difficult to handle in real-time reducedon computational Besides,conditions, spectral signatures can undergo through variationswith depending light exposureresources. and atmospheric signatures can undergo through variations depending on light exposure and atmospheric conditions, which is an issue that has been leading the scientific community to propose processes for acquisition which is an issue that has been leading the scientific community to propose processes acquisition (to control environmental conditions) and/or analysis methodologies (to correct thefor noise resulting (to control environmental conditions) and/or analysis methodologies (to correct the noise resulting from environmental conditions). Such efforts allow accurately matching spectral data and identifying from environmental conditions). Such efforts allow accurately matching spectral data and identifying material compositions. material compositions. The first developments of imaging spectrometry in remote sensing applications started using The first developments of support imagingLandsat-1 spectrometry remotethrough sensingfield applications started using satellites, more specifically to datain analysis spectral measurements, satellites, more specifically support Landsat-1 datatoanalysis field spectral according to [19]. Studiestowere mostly regarded mineralthrough exploration [20], butmeasurements, also landmine according to [19]. Studies were mostly regarded to mineral exploration [20], but also landmine detection [21], agroforestry and related areas [22]. Back then, hyperspectral imaging technology did detection [21], agroforestry and related areas [22]. Back then, hyperspectral imaging technology did not have the supporting resources to go mainstream because developments in electronics, computing not have the supporting resources to goProgress mainstream because developments electronics,technological computing and software areas were required. in the 1980s ended up toin overcome and software areas were required. Progress in the 1980s ended up to overcome technological limitations, limitations, opening the doors for the dissemination of this remote sensing technique for earth opening the doors for the dissemination of this remote sensing technique for earth monitoring by monitoring by the 1990s [23]. However, its development is still considered an ongoing process [24]. the 1990s [23]. However, its development is still considered an ongoing process [24]. Currently, Currently, satellite capabilities for wide spatial range covering along with the improvements that satellite capabilities spatial range along resolution with the improvements thatallowing have been have been carriedfor outwide regarding spatialcovering and spectral [25] have been the carried out regarding spatial and spectral resolution [25] have been allowing the development of development of remote sensing works in—also but not limited to—agriculture, forestry and related remote works in—also but not limited forestrytoand related areasraises (e.g., [26–28]). areas sensing (e.g., [26–28]). Notwithstanding, longto—agriculture, distances relatively earth surface recurrent Notwithstanding, long distances relatively to earth surface raises recurrent problems, according to [29]. problems, according to [29]. For example, Pölönen et al. [30] pointed out that for conditions involving For example, and Pölönen et growing al. [30] pointed out that for conditions involvingfrom cloudiness and short growing cloudiness short season, hyperspectral data acquired traditional platforms can season, hyperspectral data acquired from traditional platforms can become useless in some cases. become useless in some cases. Other issues include the high cost of commercial satellite imagery, Other issues of commercial imagery, which such only as provide up2toprovides 30 cm which only include providethe up high to 30 cost cm resolution [31]. satellite Even recent technology Sentinel resolution [31]. Even recent technology such as Sentinel 2 provides up to 10 m resolution in RGB and up to 10 m resolution in RGB and NIR [32], which is too coarse for some applications. For a better NIR [32], which is too coarse for some applications. For a better insight, considering a scenario with insight, considering a scenario with vines in which consecutive rows are parted by 4 m, such imagery would mix, at least, 2 rows with a significant portion of soil. An alternative to satellites started to be designed by National Aeronautics and Space Administration Jet Propulsion Laboratory (NASA/JPL) in 1983, with the development of hyperspectral hardware, specific for aircrafts, resulting in an

Remote Sens. 2017, 9, 1110

4 of 30

vines in which consecutive rows are parted by 4 m, such imagery would mix, at least, 2 rows with a significant portion of soil. An alternative to satellites started to be designed by National Aeronautics and Space Administration Jet Propulsion Laboratory (NASA/JPL) in 1983, with the development of hyperspectral hardware, specific for aircrafts, resulting in an Airborne Imaging Spectrometer (AIS). Later, in 1987, the airborne visible/infrared imaging spectrometer (AVIRIS) [33] came out as a high quality hyperspectral data provider that became popular among the scientific community [19]. However, besides the costs involved in the use of this solution, a certified pilot for manning the aerial vehicle and flight-related logistics is required. Lately, a remote sensing platform capable of overcoming not only satellite but also manned aircraft issues by bringing enhanced spectral and spatial resolutions, operational flexibility and affordability to the users is emerging: the UAS [34]. Together with specialized sensors, UAS are becoming powerful sensing systems [35] that complement the traditional sensing techniques rather than competing with them [1]. According to Pádua et al. [1], a UAS can be defined as a power-driven and reusable aircraft, operated without a human pilot on board [36]. Usually, it is composed of a UAV that, in turn, is capable of carrying remote sensing devices. UAS can be remotely controlled or have a programmed route to perform an autonomous flight using the embedded autopilot. Generally, it also requires a ground-control station and communication devices for carrying out flight missions [37]. Colomina and Molina [38] share a similar perspective by referring that a UAV is usually referred to as the remotely piloted platform, whereas the UAS is regarded as the platform and control segment. They also add that UAV and UAS are somewhat used interchangeably too. In what regards to hyperspectral data handling, a set of steps can be followed [13]: (1) image acquisition; (2) calibration; (3) spectral/spatial processing; (4) dimensionality reduction and; finally (5) computation related tasks (e.g., analysis, classification, detection, etc.). Similarly, in [39], file reduction and subsetting, spectral library definition (e.g., made by selecting a portion in the image) and classification are pointed out as valid operations to constitute a chain. Remote sensing, through the combination of UAV and on-board hyperspectral sensors and relying in many of the aforementioned steps/operations, has been applied both to agriculture and forestry (e.g., [40–42]). However, available works are not so numerous when compared with other platforms, since this is a relatively new research field. Even so, they provide a proper demonstration of this approach’s potential. All in all, the main focus of this paper is, precisely, UAS-based remote sensing using hyperspectral sensors, applied both in agriculture and forestry. The acquisition equipment designed to be attached to UAVs is presented next, in Section 2. Then, Section 3 provides a discussion towards the operations that should be carried out before and after flight missions, as well as pre-processing data procedures for image calibration. Important approaches for data processing are reviewed in Section 4, specifically data dimension, target detection, classification and vegetation indices operations. Supporting software tools and libraries are presented in Section 5. Applications focusing the use of UAV’s and hyperspectral sensors in agriculture and forestry are addressed in Section 6, right before some conclusions within Section 7. To provide guidance along this paper’s reading, a glossary regarding the used abbreviations and acronyms is listed in Appendix A. 2. Hyperspectral Sensors Independent of the aerial platform (Airborne, Satellite, UAV, etc.), sensors play an important role in data acquisition. According to [43], in which an extensive work addressing hyperspectral technology can be found, there are four main techniques for acquiring measurable data from a given target: by hyperspectral imaging, multispectral imaging, spectroscopy and RGB imagery. The most significant differences are synthetized in Table 1, which considers not only the comparison carried out by [43] but also the vision of Sellar and Boreman [44], who stated that imaging sensors for remote sensing can be divided into the method by which they achieve (1) spatial discrimination and (2) spectral discrimination. When compared with others, hyperspectral imaging sensors are effectively capable of capturing more detail in both spectral and spatial ranges. RGB imaging does not provide spectral information

Remote Sens. 2017, 9, 1110

5 of 30

beyond the visible spectrum, which is of high importance for characterizing chemical and physical proprieties of a specimen. On the other hand, spectroscopy is a proximity technology mainly used for sensing tiny areas (e.g., leaf spots), aiming the acquisition of spectral samples without spatial definition. Remote Sens. 2017, 9, 1110 5 of 31 Regarding multispectral imaging and in spite of its capability for sensing both spectral and spatial and data, there a lack of spectral resolution is only surpassed by hyperspectral data,spatial there is a lack of is spectral resolution that is onlythat surpassed by hyperspectral imaging, imaging, as it was as it wasout pointed in the section. previousThereby, section. hyperspectral Thereby, hyperspectral sensing technology be pointed in theout previous sensing technology should beshould preferred preferred whentoitsense comeschemical to senseand chemical andproperties physical properties of a specimen. when it comes physical of a specimen. Table 1. Main between hyperspectral andand multispectral imaging, spectroscopy and RGB Maindifferences differences between hyperspectral multispectral imaging, spectroscopy and RGB imagery (merging perspectives of [43,44]). A classification based on arate bullet rate (1–3) was imagery (merging perspectives of [43,44]). A classification based on a bullet (1–3) was used to used to quantify the and spectral and spatial information to each acquisition quantify both the both spectral spatial information associatedassociated to each acquisition technique, technique, in relative in relative terms. terms.

Hyperspectral Imaging Hyperspectral Imaging Multispectral Imaging Multispectral Imaging Spectroscopy Spectroscopy RGB Imagery RGB Imagery

Spectral Information Spectral Information ••• ••• •• •• ••• ••• • •

Spatial SpatialInformation Information ••• ••• ••• ••• •• ••• •••

In what what regards regardsto tothe theconcept conceptofofhyperspectral hyperspectral sensors, there area detectors In sensors, there areare area detectors thatthat havehave the the ability of quantifying acquired light that results from the conversion of incident photons into ability of quantifying acquired light that results from the conversion of incident photons into electrons [43]. [43]. Two Two types types of of sensors sensors are are prominently prominently used used to to achieve achieve such such conversion: conversion: charge-coupled charge-coupled electrons device (CCD) (CCD) and (CMOS) sensors. sensors. Both Both consist consist in in an an device and complementary complementary metal-oxide-semiconductor metal-oxide-semiconductor (CMOS) array of photodiodes that might be built using different materials, as it is detailed in Figure 3. array of photodiodes that might be built using different materials, as it is detailed in Figure 3.

Figure 3. Materials involved in hyperspectral sensors fabrication (inspired by [43]): Silicon (Si) is used Figure 3. Materials involved in hyperspectral sensors fabrication (inspired by [43]): Silicon (Si) for acquiring ultraviolet, visible and shortwave NIR regions; indium arsenide (InAs) and gallium is used for acquiring ultraviolet, visible and shortwave NIR regions; indium arsenide (InAs) and arsenide (GaAs) have a spectral between 900–1700 nm; indium gallium arsenide (GaAs) have a response spectral response between 900–1700 nm;gallium indiumarsenide gallium(InGaAs) arsenide extends the previous range to 2600 nm; and mercury cadmium tellurium (MCT or HgCdTe) (InGaAs) extends the previous range to 2600 nm; and mercury cadmium tellurium (MCT or HgCdTe) is is characterized byaalarge large spectral range quantum efficiency that enables reaching midcharacterized by spectral range andand highhigh quantum efficiency that enables reaching mid-infrared infrared region2500 (about 2500 to 25,000 NIR region (about 800–2500 nm). region (about to 25,000 nm) and nm) NIR and region (about 800–2500 nm).

CCD and CMOS sensors are different mainly in the way they treat incoming energy. On the one CCD and CMOS sensors are different mainly in the way they treat incoming energy. On the one hand, CCD requires moving the electric charges accumulated in the photodiodes into another place hand, CCD requires moving the electric charges accumulated in the photodiodes into another place wherein the quantity of charges can be measured. On the other hand, CMOS holds the photodetector wherein the quantity of charges can be measured. On the other hand, CMOS holds the photodetector and readout amplifier integrated as a single part capable of converting the voltage signal resulted and readout amplifier integrated as a single part capable of converting the voltage signal resulted from incoming electrons—converted photons—due to optically intensive transistors placed from incoming electrons—converted photons—due to optically intensive transistors placed adjacently adjacently to photodiode. This seems to be the reason why CMOS technology is faster when acquiring to photodiode. This seems to be the reason why CMOS technology is faster when acquiring and and measuring light intensity. However, it is more prone to noise and dark currents than CCD due to the on-chip circuits used to transfer and amplify signals and as a result of lower dynamic range and sensitivity, as it is explained by [43]. A dark current is a temperature-dependent common phenomenon that contributes to introduce noise in a sensor’s reading and that needs to be considered in calibration tasks, for correction purposes. According to [45], such phenomenon can be generated by the Shockley–Read–Hall (SRH) process—that considers band gap by an impurity in the lattice to

Remote Sens. 2017, 9, 1110

6 of 30

measuring light intensity. However, it is more prone to noise and dark currents than CCD due to the on-chip circuits used to transfer and amplify signals and as a result of lower dynamic range and sensitivity, as it is explained by [43]. A dark current is a temperature-dependent common phenomenon that contributes to introduce noise in a sensor’s reading and that needs to be considered in calibration tasks, for correction purposes. According to [45], such phenomenon can be generated Remote Sens. 2017, 9, 1110 6 of 31 by the Shockley–Read–Hall (SRH) process—that considers band gap by an impurity in the lattice to make make the the electron electron in in transition transition pass pass through through aa new new energy energy state—due state—due to to multiple multiple factors, factors, which which end end up resulting in the so-called blemish pixels. Additional information can be found in [15], in which up resulting in the so-called blemish pixels. Additional information can be found in [15], in which is is pointed pointed out out that that CCD-based CCD-based sensors sensors have have higher higher sensitivity sensitivity regarding regarding band band data data acquisition acquisition while, while, on on the the other other hand, hand, high high grade grade CMOS CMOS have have greater greaterquantum quantumefficiency efficiencyin inNIR. NIR. Regarding acquisition modes, reference [43] categorizes them in four main Regarding acquisition modes, reference [43] categorizes them in four main ones ones (in (in fairly fairly enough enough accordance with [44,46,47]): point scanning (or whiskbroom), line scanning (or pushbroom), accordance with [44,46,47]): point scanning (or whiskbroom), line scanning (or pushbroom), plan plan scanning scanning and and single single shot shot (Figure (Figure 4). 4). While While whiskbroom whiskbroom mode mode acquires acquires all all the the bands bands pixel pixel by by pixel pixel by by moving the detector in the x-y space to store data in a band-interleaved-by-pixel (BIP) cube, pushbroom moving the detector in the x-y space to store data in a band-interleaved-by-pixel (BIP) cube, mode proceeds similarly but,similarly instead ofbut, pixel-based scanning, an entire sequence of pixels forming pushbroom mode proceeds instead of pixel-based scanning, an entire sequence of apixels line isforming acquired, which ends up bywhich constituting (BIL) cube. Some other a line is acquired, ends upa band-interleaved-by-line by constituting a band-interleaved-by-line (BIL) pushbroom characteristics include compactinclude size, low weight, simpler operation and operation higher signal cube. Some other pushbroom characteristics compact size, low weight, simpler and to noise ratio [10]. More comparisons between pushbroom and whiskbroom modes are higher signal to noise ratio [10]. More comparisons between pushbroom and whiskbroom presented modes are in [48]. Plane mode buildsmode a band sequential cube constituted several images taken presented in scanning [48]. Plane scanning builds a band(BSQ) sequential (BSQ) cubebyconstituted by several at a time, each one holding spectral data regarding a whole given x-y space. Finally, there is a more images taken at a time, each one holding spectral data regarding a whole given x-y space. Finally, recent that acquires allthat of the spatial spectral once known single shot.asInsingle [46], there ismode a more recent mode acquires alland of the spatialdata andatspectral data atas once known snapshot imager seems to be related with the referenced single shot mode inasmuch as it is presented shot. In [46], snapshot imager seems to be related with the referenced single shot mode inasmuch as as a device that as collects an entire cube single integration some it is presented a device that data collects anwithin entireadata cube within aperiod. single Additionally, integration period. noteworthy issues are pointed outissues for each mode. Whiskbroom is mode. a slow acquisition mode Additionally, some noteworthy areacquisition pointed out for each acquisition Whiskbroom is a and pushbroom must use short enough time exposure to avoid the risk of inconsistencies at the spectral slow acquisition mode and pushbroom must use short enough time exposure to avoid the risk of band level (saturation or underexposure). Plane scanning is not suitable for moving inconsistencies at the spectral band level (saturation or underexposure). Plane scanningenvironments, is not suitable while single shot was reported as an under development technology that still lacks support to higher for moving environments, while single shot was reported as an under development technology that spatial resolutions. still lacks support to higher spatial resolutions.

Figure 4. 4. Hyperspectral Hyperspectral data data acquisition acquisition modes: modes: (A) point scanning mode; Figure (A) represents represents point scanning or or whiskbroom whiskbroom mode; in (B) presents line scanning or pushbroom mode; (C,D) correspond to plane (or area) scanning and in (B) presents line scanning or pushbroom mode; (C,D) correspond to plane (or area) scanning and single shot modes, respectively (adapted from [43]). single shot modes, respectively (adapted from [43]).

The combination of hyperspectral sensors with UAVs is commonly made available through preThe combination of hyperspectral sensors with UAVs is commonly made available through built systems that require dealing at least with three parties: the sensor manufacturer, the UAV pre-built systems that require dealing at least with three parties: the sensor manufacturer, the UAV manufacturer and the party that provides system integration [15]. The list of available commercial manufacturer and the party that provides system integration [15]. The list of available commercial hyperspectral sensors is presented in Table 2. Beside market options, others were developed in hyperspectral sensors is presented in Table 2. Beside market options, others were developed in research research initiatives (or projects). For example, in [49] a hyperspectral sensor was developed weighing initiatives (or projects). For example, in [49] a hyperspectral sensor was developed weighing 960 g, 960 g, with support for capturing 324 spectral bands (or half in the binned mode) between 361 and 961 nm. In [50], another sensor was proposed to deal with rice paddies cultivated under water. It weighs 400 g and has a capturing range of 256 bands between 340 and 763 nm. In [51], a whiskbroom imager based on a polygon mirror and compact spectrometers with a promising low-cost applicability is presented. Sellar and Boreman [44] proposed a windowing approach, distinct from the time-delay integration technique used by some panchromatic imagers. Fabry-Perot

Remote Sens. 2017, 9, 1110

7 of 30

with support for capturing 324 spectral bands (or half in the binned mode) between 361 and 961 nm. In [50], another sensor was proposed to deal with rice paddies cultivated under water. It weighs 400 g and has a capturing range of 256 bands between 340 and 763 nm. In [51], a whiskbroom imager based on a polygon mirror and compact spectrometers with a promising low-cost applicability is presented. Sellar and Boreman [44] proposed a windowing approach, distinct from the time-delay integration technique used by some panchromatic imagers. Fabry-Perot interferometer (FPI) hyperspectral imager was alternatively developed by [30] as a Raman spectroscopy device [52] behaving like a single shot imager, since it acquires the whole 2D-plane at once. The data processing steps related with this sensor are described in [53]. Pre- and post-flight operations are required to reach level-one data, i.e., spatially and spectrally reliable hyperspectral cubes, ready to use and process. Next section is devoted to such operations. Table 2. List of hyperspectral sensors (and respective characteristics) available for being coupled with UAVs. Manuf.

Sensor

Spectral Range (nm)

No. Bands

Spectral Resol. (nm)

Spatial Resol. (px)

BaySpec

OCI-UAV-1000

600–1000

100