Post on 27-Jan-2022
100s of Bands
VNIR SWIR LWIR Broadband
Hyperspectral
Hyperspectral Imaging
Why hyperspectral ? Broadband imagers (RGBI cameras,
thermal sensors, multispectral sensors) tell
you if something is Red or Green, warmer
than background, or has certain intensity
coming from certain (broad) spectral band
Hyperspectral imagers tell you `what is the
material at this pixel´
(AisaDUAL)
Push-broom hyperspectral camera
Push-broom hyperspectral imaging
Single Pixel
Spectral Bands
Spatial
Pixels
Flight
Line
Wavelength
Inte
nsi
ty Pixel
Spectrum
Single Sensor Frame
Series of Sensor Frames
Forestry • Forest inventory: ability
to classify tree stands within a forest. Precise classification of species. Biomass, leaf index, carbon cycle related measurements
• Stress detection: disease, pest, drought
• Fire science: fuel estimations, risk analysis
Courtesy of
1
2
3
4
5
6
7
8
9
11
10
12
13
14
15
16
Vegetation classification 1/2
Vegetation classification 2/2
•Vegetation classification: Invasive species mapping, vegetation community
condition and characterization, environmental baselines.
Forest chemistry University of Victoria, Canada
AisaDUAL
Example: Invasive species, Mountain pine beetle.
British Columbia, Canada
It may be the largest forest insect blight ever seen in North America. Climate change is said by some to have
contributed to the size and severity of the outbreak, and the outbreak itself may, with similar infestations, have
significant effects on the capability of northern forests to remove greenhouse gas (CO2) from the atmosphere.
16
Emerald Ash Borer Detection
The emerald ash borer is now one of the most destructive non-native insects in the United States;
it and other wood-boring pests cause an estimated $3.5 billion in annual damages in the U.S.[
Citrus canker detection
Leaf infected with citrus
canker.
Aerial photo of
citrus grove from
3,000 feet.
Hyperspectral image processed to highlight
canker infestations in red.
Courtesy of
Citrus canker is a disease affecting citrus species that is caused by the
bacterium Xanthomonas axonopodis. Infection causes lesions on the leaves,
stems, and fruit of citrus trees, including lime, oranges, and grapefruit. While not
harmful to humans, canker significantly affects the vitality of citrus trees, causing
leaves and fruit to drop prematurely
Fungus disease in oil palm trees
Airborne HSI in VNIR provides sensitive and high resolution detection and mapping of a fungus disease in oil palm trees
>50 km2/h @0.5 m ground resolution @50 m/s (100 knots)
Sarawak Forest Department, Malaysia
Wetlands Management – Phragmities
Wetlands Management – Phragmities
Biomass mapping
• Goal: monitoring crop vigor and disease treatment results in
high-value crops.
Green areas show
the improved biomass of crops after applying fungicide toxicant.
Courtesy of
High
Low
Test Field Mapping
• Agricultural Applications
– Differentiation of various types of corn
– Extent of drought
– Mapping of economically important vegetation/illegal logging
– Detection of spectrally coded, genetically manipulated plants
– Socioeconomic assistance
Agricultural/Vegetation Mapping
Fire Fuel Modeling – Orange County, CA
• Invasive/Target Species
– Identify eucalyptus (in BLUE) in fire
district
– Analysis: Utilize
spectral signature or
(fingerprint) of species
Invasive/Target Species
•Asphalt roads were masked out to eliminate
confusion between Asphalt shingles.
• Spectrally able to identify every
house in the image and
determine its material
composition. Ground Verification
done and determined 100%
accuracy.
Roof Material ID
Fire Risk Assessment
Rooftop classification and fuels mapping can be done to identify
homes at greatest risk in the event of a wildfire
Contra Costa County, CA
Land Cover Classification Applied to Fire
Management by University of Zurich (1/2)
• Goal: Improved risk
assessment and
migitation of forest fires
• LIDAR system and AISA
Eagle were used to study
mediterranean
vegetation intermixed
with urban structures
• The joint classification of
AISA Eagle and LIDAR led
to significant
improvements in terms
of overall accuracy and
kappa
Inversion
Hyperspectral Reflectance
Derived Bathymetry
Near Port Hedland,
Western Australia
September 2010
Collaborators:
• WAMSI (WA Marine
Sciences Institute)
• ARA - Flinders U
• Curtin University
• In Situ Marine Optics
FLINDERS UNIVERSITY ADELAIDE ● AUSTRALIA
Depth (m)
0 5 10
Water quality mapping
Chlorophyll-a map Phycocyanin map Total solids content
University of Nevada, USA
Quality monitoring and
mapping in
- Water reservoirs
- Lakes, rivers
- Coastal areas
Hydrological Studies and Risk Assesment • The town of Szeged,
Hungary, has been several times flooded by the river Tisza. OGS utilized AISA Eagle hyperspectral sensor to carry out a research on the area.
• Top: Land cover map, valuable in the computation of a hydraulic model
• Normalized Difference Vegetation Index (NDVI) can be applied to estimation of water content in the terrain and assisting in the computation of risk maps
Courtesy of
Chlorophyll ConcentrationHood Canal, Washington
Chlorophyll-a μg/l
10
35+
Water TurbidityHood Canal, Washington
Relative Turbidity
Low
High
Water Quality - Agricultural Runoff
Geologic Applications
Geothermal Exploration: Subtle geothermal related
alterations of associated minerals such as Illites
Mineral Precipitates as Acid
Mine Drainage Indicators Courtesy of D. Coulter – Overhill Imaging Ltd.
SpecTIR 1m GSD (400nm – 2500nm)
Environmental catastrophes
Oil spill in the Gulf of Mexico 2010:
Hyperspectral data is frequently collected with
AisaDUAL by SpecTIR LLC, USA, in order to
- monitor changes to health in highly sensitive
coastal wetlands, and
- aid in future impact assessments.
‘Red mud’ flood in Hungary 2010:
Hyperspectral data is collected with AisaEAGLE
by EnviroSense , Hungary.
Eutrophication Mapping of Water Bodies
Direct Hydrocarbon Detection
Aerial photograph of an oil spill in Buzzards Bay,
MA. Red areas are an interpretet’s estimates of
oil extent in the water. Blue arrows added by the
author point to areas that are wet where it is
unclear whether water or oil is the wetting agent
(Photo source: Buzzards Bay, 2003)
AISA DUAL False Color Composite of some
POL tanks 1 week after Katrina. Location is
near New Orleans, LA. It is unclear from
literal interpretation whether the areas inside
the berms are covered in water or petroleum.
Petrole
um
Oil spill monitoring
• Rapid and accurate assessment of the damage helps to maximize the cleanup efforts.
• Environmentally sensitive areas can be targeted for protection and cleanup, and the long-term damage can be minimized.
• Time sequence images of the oil can guide efforts in real-time by providing relative concentrations and accurate locations.
Courtesy of
The smallest and lightest Hyperspectral LWIR
imager available
Several customers have described this as ”best
thermal spectral data they ever have seen”
AisaOWL (7.6-12.5µm)
AisaOWL- methane gas leak detection
Outdoor scan in Finland in daytime in February 2012
Ambient temperature ca -5 C
Methane flow rate 20 l/min.
Methane absorption around 7.7 um.
Outside and gas temperature about -15C
AisaOWL scan
Detection of camouflaged targets
Queries or Enquires ??
Please contact Mr. Venugopal Parasuraman
1342 A, Road No 67 , Jubilee Hills, Hyderabad 500033.
Phone : +91 - 40 - 6460 - 4343
Fax : +91 - 40 - 2355 – 2966
Mobile : +91 - 95021 - 65836
Email : Venugopal.Parasuraman@AmigoOptima.com
www.amigooptima.com