See more with hyperspectral imaging

See more with hyperspectral

imaging

SPECIM, Spectral Imaging Ltd.

World leading manufacturer and suppplier for

hyperspectral imaging technology and solutions

Hundreds of customers worldwide.

Distributor and integrator network covers more

than 40 countries

We make hyperspectral imaging possible

What is Hyperspectral Imaging?

• Each pixel is associated with the spectral signature of the target

• Spectral signature includes information about - precise color- chemical composition, and/or- temperature

• Hyperspectral imaging is used in new generation- machine vision solutions- life science imaging systems- QA and process control systems- airborne and security imagers

R

400 700

Sp

ectr

al

axis

Spatial axis

Tim

e

Wavelength

Spectrum for one pixel

Simultaneous acquisition of data from hundreds of

contiguous spectral bands full spectral and spatial

information for each pixel

Visible

SWIR LWIR

Broadband

Hyperspectral

Hyperspectral Imaging

LWIR

VNIR

Visible

VNIR

SWIR

SWIR

Multispectral and Hyperspectral Imaging

SPECIM Push-broom Hyperspectral Camera

Imaging Spectrograph

disperses light to its

spectral elements

Spectral Camera =

Lens + Spectrograph + Camera

Products and customers

Hyperspectral CoresImSpector imaging spectrographs

Spectral Cameras

SISU Chemical

Imaging Workstations

AISA Airborne

Hyperspectral

Systems

Instrument manufacturers

System integrators

Science customers

Pharma and agri-food

industries

Mining industry

Forensics labs

Science organisations

Remote sensing service

providers

Security and defense org.

AISA

Airborne Hyperspectral Systems

Single Pixel

Spectral Bands

SpatialPixels

Flight

Line

Wavelength

Inte

nsi

ty

Pixel Spectrum

Single Sensor Frame

Series of Sensor Frames

Airborne imaging: Push-broom Hyperspectral Sensors

AisaEAGLE

• Spectral range: 400 – 970 nm

• Spectral resolution: 3.3 nm

• Spectral bands: 488

• Spatial pixels: 1 024

• Light and compact

VNIR Hyperspectral Sensor

APPLICATIONS

Vegetation, farming

Environmental monitoring

Defense & Security

AisaEAGLET

• Spectral range: 400 – 1000 nm

• Spectral resolution: 3.3 nm


• Spatial pixels: 1 600

• Exceptionally light and compact

VNIR Hyperspectral Sensor for small aircrafts or UAVs

APPLICATIONS

Vegetation, farming


Defense & Security

AisaHAWK

• Spectral range: 970 – 2 500 nm

• Spectral resolution: 12 nm


• Spatial pixels: 320

SWIR Hyperspectral Sensor

APPLICATIONS

Geology,minearology


Defense & Security

AisaDUAL

APPLICATIONS

Vegetation

Geology,minearology


Defense & Security

Simultaneous VNIR-SWIR data collection

Overview of the

Goldfield and

Cuprite alteration

systems.

High spatial resolution

(1 meter) hyperspectral

results for the “Buddintonite

Bump” area of Cuprite,

Nevada. Image courtesy of

SpecTIR LCC

• Combining AisaEAGLE & AisaHAWK

• Nearly perfect alignment of VNIR and SWIR

• Single camera port needed

• Light and compact

AisaOWL

Hyperspectral Imaging in LWIR

Applications:

– Geological mapping

– Mineral classification

– Volcanology

– Ground and coastal water temperature

– Camouflage detection

– Gas detection

– Flame analysis

– Land cover type recognition

• Spectral range: 8 – 12 um

• Spectral resolution: 100 nm


• Spatial pixels: 384

AISA Airborne Hyperspectral

System: Components

Data acquisition and

storage system in a

rugged industrial PC

Hyperspectral

sensor

GPS/INS unit

Daylight readable

LCD DisplayFODIS

CaliGeo Software

RSCube Software

• Control setup options (width, position and number of spectral bands)

• Control hardware (image rate and exposure time)

• Displays images, GPS/INS status, and other information in real-time for monitoring the progress of data collection

• 2 operating modes:

– Full hyperspectral data acquisition

– Multispectral data acquisition at programmable wavebands

Flight Operations Software

CaliGeo Software

• User-friendly, ENVI

compatible

– Radiometric correction

– Geo-referencing

– Provides the tools to

automatically correct sensor vs.

GPS/INS unit misalignment

Advanced AISA pre-processing software

Current Owners of AISA systems in Operational Use

North America

US:

Galileo Group

Helicopter Applicators

University of Nebraska

Texas A&M University

University of Montana

R&D Organization

University of Indiana

SpecTIR

R&D Organization

University of Cincinnati

Canada:

University of Victoria

In total more than 60 sensors delivered

Europe

UK: Infoterra/Astrium, R&D Organization, Natural Environment Research Council

(NERC), R&D organization

Hungary: University of Debrecen, University of Gödöllö

France: University of Rennes

Germany: Anhalt University, Alfred Wegener Institute, Helmholtz-Zentrum für

Umweltforschung GmbH – UFZ, FGAN-FOM

Italy: OGS

Czech Republic: Academy of Sciences

Turkey: R&D Organization

Finland: Forest Research Institute, Helsinki University of Technology

Asia

Japan: Pasco Corp., R&D Organization

China: State Oceanographic Agency, R&D institute

India: R&D organization

Malaysia: University Putra Malaysia, Aeroscan, Forest Dept. of Sarawak

Australia

ARA ( Airborne Research Australia)

AISA applications

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

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

Farm mappingHigh

Low

Unsupervised classification Supervised classification Relative biomass

Courtesy of

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

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

Land Cover Classification Applied to Fire

Management by University of Zurich (2/2)

Fusion of LIDAR and AISA Data in Forest Study by the

University of Victoria, Canada (1/2)

• A Digital Surface Model (DSM) model is created

from LiDAR data. DSM is resampled to same

resolution with hyperspectral image and used

for orthorectification of hyperspectral AISA

data.

• The second step of LiDAR data processing is

vegetation ground classification. For both

visualization and analysis purposes the

hyperspectral data is overlaid on surface

model created from LiDAR data.

• Eventually targets are classified using for

example Spectral Angle Mapper replenished

with some other sophisticated methods

• Tree lists with

associated spectra

for each tree have

been developed.

These spectra are

being used to assign

species to the

individual trees Small sample of the LIDAR canopy

height model with tree topsAISA data with tree tops as defined by the

canopy height model.

Fusion of LIDAR and AISA Data in Forest Study

by the University of Victoria, Canada (2/2)

Mineral Exploration

• The image maps alunite, kaolinite, and illite as RGB. Data where acquired by Spectir LLC Using a AISA Dual imaging spectrometer.

• The results illustrate the advantage of high spectral resolution hyperspectral imagery for mapping alteration minerals at the outcrop scale. Overview of the

Goldfield and Cuprite

alteration systems.

High spatial resolution (1 meter)

hyperspectral results for

the “Buddintonite Bump” area of

Cuprite, Nevada.

Automatic Classification of Sedimentary

Stratigraphy (Ragona et al.)

• Ground-based hyperspectral imaging enables the study and digitally storing of stratigraphic and structural data from samples in the lab (cut samples, drill cores) or field (outcrops, exposures).

• both a large sample of 1m x 1m and several drill cores were used to evaluate the work flow and classification results.

Coral Reef Mapping

• Right: false colour composite. The limitations of 3 band displays make it difficult to demonstrate the depth of spectral information.

• Middle: Composite of 3 MNF (Minimum Noise Fraction) images that were calculated from all the 29 available spectral bands. MNF compresses the amount of coherent signal found in the complete data set to just a few derived bands.

• Left: The scope of information within the imagery as can be seen by the simple application of a colour ramp to a single MNF band.

Courtesy of

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

Water quality studies

Unsupervised classification

Chlorophyll-a map Phycocyanin map Total solids content

Images courtesy of

University of Nebraska

Bathymetry of St.

Kilda Channel,

Adelaide

Measured with

AisaEAGLE

True color RGB

Bathymetry

Images courtesy of ARA

(Airborne Research

Australia)

Oil palm mapping

Hyperspectral data collected

with AisaEAGLE

Palm tree map created with

palm tree mapping algorithm

Images courtesy of

Forestry Department of

Sarawak State in Malaysia

Map of infected oil palms

(red indicates healthy trees,

yellow infected ones)

Your partner in

hyperspectral imaging

See more with hyperspectral imaging

Documents

Transcript of See more with hyperspectral imaging