Codevintec

Report
Innovative applications of
Specim's AISA hyperspectral
sensors and software
Ing. Marco Bacciocchi, Codevintec
Hyperspectral sensors and software
Current Airborne AISA Products
AISA Systems
EAGLE/EAGLET
HAWK
DUAL
OWL
Spectral range
400-1000 nm
970-2500 nm
400-2500 nm
7.7-12.4 um
Spectral sampl.
3/2.5 nm
6 nm
2.5/6 nm
50 nm
Spectral bands
200/244
256
498
100
Spatial pixels
1024/1600
320
320
384
Image rate
125/160 Hz
100 Hz
100 Hz
100 Hz
SNR
130-1400
800
Sensor weight
3.5/6.5 kg
18 kg
55 kg
13.5 kg
Total weight
15/28 kg
45 kg
110 kg
35 kg
250-500
AISA Airborne Hyperspectral
System Components
Daylight readable
LCD Display
Hyperspectral
camera
FODIS
CaliGeo Software
GPS/INS unit
Compact Data acquisition
and storage system
AISA Installations
Push-broom hyperspectral imaging
Intensity
Flight
Line
Single Pixel
Pixel
Spectrum
Wavelength
Spatial
Pixels
Spectral Bands
Photo - SCG Siena
Single Sensor Frame
Series of Sensor Frames
Push-broom hyperspectral camera
Push-broom advantages
1.
Acquires all spectral information exactly at the same time - insensitive to
instrument/sample movement
2.
Multiplex advantage of imaging full line of pixels at the same time
3.
No moving parts in the instrument – compact, reliable, stable, low maintenance.
4.
With high speed AISA sensors, flexible to adapt to various mission requirements
5.
The only HSI technique which practically fits to all
applications: air, field, lab and on-line
Flight mission parameters
GSD = ground sampling [m]
v = aircraft speed [m/s}
FOV = field of view [degree]
n = # of swath pixels
Frame rate FR = v/GSD [1/s]
Flying altitude
A = n*GSD/2tan(FOV/2) [m]
Frame integration time [s]
AISA system lab calibration
Radiometric & spectral
› Conversion factor for each pixel from raw
data to spectral radiance
› Centre wavelength and spectral FWHM for
each spectral band
Geometric
› Focal length
› Optical axis position
› Possible distortion over the swath
Focus
Other standard lab tests
› Linearity, Signal-to-Noise Ratio, Smile and Keystone, Stability, MTF
Data processing
Radiometric processing
Georeferencing
Atmospheric correction
Mosaic of flight lines
Spectral processing for desired information
CaligeoPRO
› Both interactive and batch mode
Atcor, Flaash
ENVI, et al
AISA software
Aisa Operating Software RS Cube
All the AISA systems use
Windows-based flight operations
software RSCube to
› Control hardware like image
(frame) rate and exposure time.
› Display images, GPS/INS status,
and other information in realtime for monitoring the progress
of data collection
Aisa Operating Software RS CUbe
All AISA systems can be used
in two operating modes:
› A: full hyperspectral data
acquisition
› B: multispectral data
acquisition at programmable
wavebands Band files are
created with AISA Bandage
Caligeo Software
› CaliGeo is an easy-to-use, GUI-based, interactive software package (stand
alone or Envi plug-in). It allows you to fine tune your images and turn your raw
AISA data into final useable products.
› With CaliGeo you are able to perform
› Radiometric correction
› Geo-referencing (Geometric correction)
› CaliGeo also provides the tools to automatically analyze and correct boresight
errors caused by mechanical tolerances between:
› head and GPS/INS
› VNIR and SWIR sensor heads
Dems
› A digital elevation model (DEM) defines the ground elevation for every single ground point,
independently of the other points.
› Using a DEM is supported in CaliGeo. However, it is up to the user to provide ENVI-readable
DEM data for the entire image area in a suitable format with a proper resolution
› If the DEM does not cover the image area and CaliGeo needs values outside the DEM, then
CaliGeo will use the nearest DEM pixel on the edge of the DEM.
› There are no size restrictions for the DEM file. The file can cover a larger area than the raw
image.
› The DEM file must be in ENVI readable format (1 channel grayscale raster data)
› To achieve a good rectification result, the ground resolution of DEM should reflect the height
differences in ground (e.g. If having tall buildings etc. in the image a good ground resolution is
recommended)
AISA Applications and References
Why hyperspectral?
Hyperspectral imaging is capable of seeing
detailed spectral signatures needed in
› identification (like plant species, minerals)
and
› quantification (like chlorophyll in water)
of target characterisctics, and
› mapping their distribution.
Where hyperspectral imaging can
help us?

Mineral Exploration & Geothermal Exploration, Acid Mine Drainage

Water Quality, Off-shore Mapping, Coral Reef Health

Forest Chemistry, Health, Inventory

Monitoring Infrastructure Conditions

Wetlands Health, Vegetation, Discrete vegetation mapping, and
Agriculture Research and Development

Hydrocarbon Detection, Oil&Gas Leak Detection

Fire and Flood Risk Analysis

Law Enforcement and Military Applications
Airborne mineral mapping
SpecTIR LLC, USA
Operates three AisaDUAL
systems, first obtained in
2006
SWIR and LWIR hyperspectral
imaging is capable of identification
of most minerals of commercial
interest, and provide quick
mapping tool in geological and
geothermal exploration.
High spatial resolution (1 meter) hyperspectral
results for the “Buddintonite Bump” area of
Cuprite, Nevada.
Airborne mineral mapping
Courtesy of SpecTIR LLC, USA
Forest inventory and health
University of Victoria, Canada
Purchased AisaDUAL in 2006
› Fusion of hyperspectral imaging and Lidar
data
› Tree height and volume
› Species identification and distribution
› Biomass, tree health condition, like stress
caused by pine beetle attack
Forest chemistry
University of Victoria, Canada
AisaDUAL
Forest health mapping
Sarawak Forest Department
Malaysia
AisaEAGLE
Airborne HSI in VNIR provides
sensitive and high resolution
detection and mapping of fungus
disease in oil palm trees
>50 km2/h
@0.5 m ground resolution
@50 m/s (100 knots)
Law enforcement in forest area
Sarawak Forest Department
Malaysia
Purchased AisaEAGLE in 2008,
AisaEAGLET in 2010
Monitoring of illegal logging and
encroachment by detection,
classification and mapping of
anomalies, like natural vs. felling
gap, non-vegetative green
targets (camouflage)
Integrated geospatial system for fast
enforcement response:
› hyperspectral data collection
› near real-time processing and mapping
› web dissemination of the tactical
information to field users
Detection of recent encroachments
Sarawak Forest Department
Malaysia
Drug plant detection
Processed Hyperspectral Image
• Known spectral characteristics
• Known background
• Geo-rectified
Marijuana
plants
Courtesy of SpecTIR LLC, USA
Water quality mapping
SOA (State Oceanic
Administration), China
Purchased AisaEAGLE 2006,
and two more systems in 2009
Chlorophyll, algae, and total
solids mapping
Oil slick and oil in water
monitoring
University of Nebraska, USA
Purchased AisaEAGLE 2005
Chlorophyll-a map
Phycocyanin map
Total solids content
Water quality mapping
Centro de Economia
Aplicada (CEA), Chile
Purchased AisaEAGLE
In 2009
All water reservoirs
in Chile are frequently
mapped for organic
and inorganic pollution
Environmental catastrophes
Oil spill in the Gulf of Mexico:
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:
Hyperspectral data is collected with AisaEAGLE
by EnviroSense , Hungary.
Data courtesy of
Thermal applications
AisaOWL Key Characteristics
285x200x175 mm
13.5 kg
< 200 W
Spatial pixels
384
Spectral range
7.7-12.4 um
Spectral resol.
100 nm
Spectral sampling
50 nm
Spectral bands
100
Smile, keystone
<0.2 pix
Image rate
100 Hz
[email protected]
20 mW/m2 sr um
Thermal Cooled – Performance
NESR
SNR
AisaOWL sensor on the ground
Detection of propellant gas (1,1,1-2 tetrafluoroethane)
Outdoor scan in Finland in daytime in March 2011.
Ambient temperature ca -10 C.
Data processed to radiance.
Codevintec Italiana
Thank you!
Via Labus 13
I-20147 Milano
Tel. +39 02 4830.2175
[email protected]
www.codevintec.it

similar documents