Home Articles Hyperspectral Remote Sensing -an Overview

Hyperspectral Remote Sensing -an Overview

Anshu Miglani
Assistant Editor
GIS Development
[email protected]

Since the time when Sir Isaac Newton published the concept of dispersion of light, the scientific terminology and definitions of the term “spectroscopy” has evolved over time. Today, imaging spectroscopy or “hyperspectral imaging” is defined by a contiguous statement of spectral bands. It is the science of acquiring digital imagery of earth materials in many spectrally narrow contiguous bands and produce complete spectral signatures with no wavelength omissions. The image produced is similar to an image produced by a multispectral sensors, except that each pixel has many bands of light intensity data instead of just three bands: red, green and blue. Hyperspectral imaging collects data to generate a “data cube” or “image cube” that can reveal objects and information which conventional multispectral scanners cannot pick up. Each element leaves a unique spectral signature, also called spectral reflectance curve, in various bands of the spectrum, based on their spectral reflectance.

The overall shape of the spectral curve and the position and strength of absorption bands is used to identify different materials. The reflectance spectra of various materials of known composition can be measured in the field or laboratory and can be collected as spectral libraries to facilitate analysis of hyperspectral imagery. Several high quality spectral libraries of reflectance spectra of natural and man-made materials are available for public use. These libraries provide a source of reference spectra that can aid the interpretation of hyperspectral and multispectral images. Various type of spectra can be browsed, viewed and downloaded from USGS . usgs.gov/spectral.lib04/spectral-lib04. html) and ASTER . nasa.gov) website.

Multispectral RS deals with several images at “discrete and narrow band”, from the visible to the Infrared wavelength, whereas hyperspectral RS deals with imaging narrow spectral bands over a contiguous spectral range, and produce the spectra of all pixels in the scene. It is not the number of measured wavelengths that defines a sensor as hyperspectral, rather it is the narrowness and contiguous nature of the measurements.

Thus, a sensor with only 20 bands can be hyperspectral when it covers the range from 500-700 nm with 20 10-nm wide bands, while a sensor with 20 discrete bands covering the VIS, NIR, SWIR, MWIR, and LWIR would be considered multispectral.

“It is not the number of measured wavelengths that defines a sensor as hyperspectral, rather it is the narrowness and contiguous nature of the measurements.

Data from hyperspectral remote sensing provides more details of the Earth’s surface than is currently available from multispectral instruments.

The detailed classification of complex land ecosystems with hyperspectral imagery is expected to increase the accuracy of remote sensing data in applications including mining, geology, forestry, agriculture and environmental management.

Projects utilizing hyperspectral imagery usually have one of the following objectives:

  • Target detection: Involves distinguishing targets from similar backgrounds, or locating examples of targets that are smaller than the nominal pixel size.
  • Material identification: The analysis is designed to use hyperspectral imagery for identifying the unknown materials. This analysis may also be accompanied with material mapping in which the identified materials are geographically located throughout the image.
  • Material differentiation: It is carried out to distinguish between spectrally similar materials.
  • Hyperspectral imagery has also been used for mapping surface properties that are undetectable using other types of imagery.

Challenges with hyperspectral imagery

Hyperspectral images contain wealth of data, but interpreting them is a very big challenge. It requires an understanding of exactly what properties of ground materials are being measured and how they are related to the measurements actually made by the hyperspectral sensor. Although the potential of hyperspectral remote sensing is exciting, the following issues needs to be considered while analysis/ processing of this unique type of imagery,

  • Accurate atmospheric corrections
  • Availability of spectral libraries
  • Spectral mixing

Hyperspectral Missions: List of space/ aircraft based sensors

Hyperspectral imagery is often not as readily available as other types of remotely sensed data. In particular, there are few space borne hyperspectral sensors, including the Hyperion sensor on NASA’s EO-1 satellite, the CHRIS sensor on the European Space Agency’s PROBA satellite and the FTHSI sensor on the U.S. Air Force Research Lab’s MightySat II satellite.

Future Hyperspectral Missions
Hyperspectral instruments will be the mantra for the next generation of optical satellite sensors. Reflection spectra from the earth’s surface give researchers a broad basis for detailed analyses and the generation of variety of data products. Several countries are striving for developing their own hyperspectral sensor for its applications in the environment monitoring, risk assessment, mineral exploration etc.

For almost two decades now, there has been worldwide research and development into the application of high spectral resolution remote sensing to various earth resource and environmental mapping and monitoring tasks. Hyperspectral image analysis has matured into one of the most powerful and fastest growing technologies in the field of remote sensing. This has the potential for more accurate and detailed information extraction than possible with any other type of remotely sensed data, available today. Extracting the reflection spectra information from myriad of bands from the earth’s surface give scientists, a broad basis for detailed analyses. For the next generation of optical satellite sensors hyperspectral instruments will play an important role.

The success of applying such techniques often relies on the detection of subtle variations in the spectral properties of one or more of the components being imaged and on the integrity of Board Band Spectrum (e.g. Landsat TM) Continuous Spectrum (e.g. Imaging spectrometer) OmniProbe, the third generation imaging instrument have been patented by the Earth Search Sciences, Inc. which is a commercial provider of hyperspectral remote sensing technology. Based on nanotechnology and Microelectronics. Its weighs less than 40 pounds and it can can attached to virtually any airborne platform. OmniProbe will quickly and accurately map targets of interest from altitudes of 35,000 feet, at speeds of 500 to 600 knots per hour, and will possess telescoping optics to scope down to 6 inch data. After the success of Probe-1, OmniProbe is sure to be a blessing for geoexploration. OmniProbe is the cutting edge imaging instrument as compared to its counter parts because of its improved resolution (both spectrally and spatially) and low cost and thus wider range of applications. According to Larry Vance, the Chairman of Earth Search, “As we move into the next stage of our strategy, our newly-advanced instrument will enable us to joint venture with mineral, hydrocarbon, environmental, and military resource industries to quickly map targets of interest more precisely and economically than any instrument with which we’re familiar.” OmniProbe: Third Generation Imaging Instrument the sensor’s calibration. Such requirements have driven sensor technology to achieve higher signal to noise ratios, improved operational stability and improved levels of traceable spectral and radiometric calibration. As new sensors provide more hyperspectral imagery and new image processing algorithms continue to be developed, hyperspectral imagery will become a powerful tool for research, exploration of minerals and monitoring of natural resources and their sustainability.