EOS 740 Hyperspectral Imaging Systems

1
EOS 740 Hyperspectral Imaging Systems
February 11, 2005 Week 3
Ron Resmini v 703-735-3899 ronald.g.resmini_at_boein
g.com Office hours by appointment
Put EOS740 in the subject line of e-mails to
me...Thanks!
2
Outline

• The nature of HSI the data cube a review
• Scientific principles of HSI RS
• The remote sensing process
• Information extraction
• Using ENVI a review
• More information on your HSI data
• A thread through the HSI analysis process
• Building a product/exporting from ENVI
• Band and spectral math
• Spectral libraries

3
Hyperspectral Sensing Concept (Cont.)
Graphic from JPL
4
Properties of the Data Cube

• of samples, lines, bands
• Headers, preline, postline, footers, etc.
• Data type
• Interleaving
• Byte order
• Center wavelengths, FWHM
• Bad bands list
• Band names (very optional)
• The logical and physical data cube
• The ENVI .hdr file
• History file (it doesnt exist) keep notes!

5
A Few Scientific Principles of HSI RS

• Sensors measure radiance as a function of
  wavelength
• Radiance (W/m2.sr.mm) (spectral)
• Materials interact with electromagnetic radiation
  (EMR)
• Materials reflect, absorb, and/or transmit EMR
• HSI is based on discerning/measuring the
  interaction oflight (photons, waves) with matter
• Other radiometric quantities, units, etc. of RS
• Irradiance (W/m2.mm) (spectral)
• Reflectance
• Emissivity

6

• Traditional HSI spectral ranges
• VNIR/SWIR (0.4 to 2.5 mm), MWIR (3 to 5 mm), LWIR
  (7 to 13 mm)
• Determined by h/w considerations and atmospheric
  windows
• Do not be so constrained when considering other
  apps. for HSI
• HSI is really a problem in inversion we sense
  the answerwe work backwards from there we
  sense boundary conditionsin one instant in time
• Key statement
• The spectrum is the fundamental datumin imaging
  spectrometry

7
The Spectrum is the Fundamental Datum in Imaging
Spectrometry
8

• What you need to know about your data a
  check-list
• Date, time, location, ground elevation, platform
  elevation,heading, GSD, be able to calculate
  where the sun isi.e., all RS angles (geometry)
• On-going sensor characterization know what it
  is ask for it!
• Spatial sampling spatial resolution
• Spectral sampling SRF spectral resolution
• NESR, NEDr, NEDe, NEDT
• Issues smile, keystone, FPA misregistration,
  vibration,parallax, scattered light,
  self-emission, platformmotion/imaging
  distortions, etc. H/W guest lecturer willcover
  these.

9
HSI RS facilitates remote material identification.
This capability also allows material
characterization and quantification. Its
spectroscopy.
10
(No Transcript)
11
(No Transcript)
12
Information Content and Extraction

• HSI RS is based on the measurement of a physical
  quantityas a function of wavelength its
  spectroscopy
• HSI is based on discerning/measuring the
  interaction oflight (photons, waves) with matter
• The sun is the source or active systems or very
  hot objects
• Earth RS scenarios involve the atmosphere
• There are complex interactions in the atmosphere
• There are complex interactions between light and
  targetsof interest in a scene
• There are complex interactions between light,
  targets ofinterest, and the atmosphere
• Theres a lot (lots!) of information in the
  spectra

13
Working with ENVI (1 of 2)

• Hows the homework assignment going?
• Data CD contents questions?
• Help files
• Open files
• Grayscale images
• Histogram/interactive stretching
• Animation
• Color composite images
• Bad bands list
• Looking at spectra radiance and reflectance
• Data statistics

14
Working with ENVI (2 of 2)

• Plot window features
• Multiple cubes linking data
• Exporting information
• Building a product
• Information extraction introduction
• Spectral matching overview
• Spectral matching with SAM
• Band and spectral math
• Spectral libraries in ENVI

15
What Were Going to Review

• Spectra as vectors points in hyperspace
• Angular separation of vectors (spectra)
• Spectral Angle Mapper (SAM)
• Invariant to albedo...wait a couple of slides
• Running SAM in ENVI
• Application strategies(i.e., in-scene
  spectra/library spectra)
• Mixed pixels...and SAM...

16
The Geometry
Angular Distance Metric (Spectral Angle Mapper or
SAM)
Assume a two band spectral remote sensing system.
Each two point spectrum is a point in Band b
vs. Band a space.
A 2D scatterplot with 2 spectra
The angle, Q, between the two lines connecting
each spectrum (point) to the origin is the
angular separation of the two spectra. Smaller
angular separations in- dicate more similar
spectra.
17
The Math

• Chang (2003), ch. 2, pp. 20-21 (see .pdf
  file) and...
• Assume two 5-band spectra as shown

BTW...read Sec. 2.2 to Sec. 2.2.2 on pp. 20-21
fair game material for the mid-term.
18

• Let the 5 bands have band names a, b, c, d, and e

• The output units are radians
• ENVI does all this for you

19

• Invariant to albedo...why

A 2D scatterplot with 2 spectra
20

• Application strategies
• A few comments on SAM andmixed pixels
• Ch. 2 in Chang (2003) can/shouldENVI do any of
  these?

21

• Band math
• Spectral math
• CBD, ED from Chang (2003)

22
Assignment 2 Due the Week of 25 Feb., 2005
Under Construction
23
Project Challenges Ive added some...

• N-P Theory sensitivity to spatial/spectral
  subsets
• When is spectral mixing linear v. non-linear?
  I.e., is this evidentfrom the spectra?
• Measure the volume of hyperspace actually
  occupied by real HSI data
• Spectral angle between spectra and filter
  vectors is the separabilitygreater than angles
  derived from a confusion matrix analysis of
  aspectral library? Use, also, a measure of SCR
• Make Mine Virginia Wine. Characterize VA
  vineyard soils with HyperionHSI and/or field
  spectrometry characterize grape vines
  etc...Does HSI have a role in the VA wine
  business?
• Test various algorithms with target-implanted HSI
  data sets
• Compare recently published TES routines
• Evaluate noise removal/compensation algorithms
• Spectral indicators for urban lead poisoning and
  medical geology

24
Project Challenges (continued)

• Invert the SAIL canopy RT model with noise...
• Implement, compare, test algorithms from the
  textbook
• Assess impact of spectral MTF on subpixel
  unmixing
• GSD and the geometry of hyperspace...
• Derivative spectroscopy and vegetation RS e.g.,
  REDE
• Optical bathymetry (e.g., swimming pools)
• Continuation of projects from last semester

BTW...Contributions to scientific knowledgevice
generic techniques studies are also strongly
encouraged.