Transcript Extending hyperspectral capabilities with

PHYSICS
Demonstration of a
Dualband IR imaging Spectrometer
27 August 2007
San Diego, CA
SPIE Conference 6660A
Infrared Detectors and Focal Plane Arrays IX
Brian P. Beecken
Physics Dept., Bethel University
Paul D. LeVan
Air Force Research Lab, Kirtland AFB
Benjamin D. Todt
Physics Dept., Bethel University
1
PHYSICS
•
•
•
Classic “2 channel” Spectrometer
Efficiencies change with λ
– Gratings
– FPA detectors
Classic Solution: 2 channels
– Common aperture & FOV
– Beamsplitter
– 2 Dispersive elements
and 2 FPAs
– Each channel optimized
for roughly 1 octave of λ
Issues
– Size
– Mass
– Power consumption
– λ Registration
– Complex
FPA
Dispersive Elements
2
PHYSICS
Dualband FPA Diffraction Concept
Spectral Image, but only 1 spatial dimension
Spatial Dimension
Dualband
FPA
Multispectral IR
Spectral
Dimension
Dispersive
Element
3
Using Dual-band FPA
PHYSICS
•
Gratings
–
–
•
1st order
is LWIR
nλ = d sin θ
2nd order is MWIR
Peak efficiencies at
λB, λB/2, λB/3,…
Designed Bands:
3.75 – 6.05 µm (MWIR)
7.5 – 12.1 µm
(LWIR)
•
λ Gap chosen to prevent
spectral crosstalk
•
Advantages:
–
–
–
–
Reduced Complexity
Smaller mass & size
Less cooling required
Perfect λ registration
320 cols x 240 rows
4
PHYSICS
Schematic of Dewar Optics
Dualband
FPA
grating
Image formed on slit
5
PHYSICS
Solar Observations
• Goal: demonstrate functionality
• Why the Sun?
– Distant
– Extended body for imaging
– Significant IR signature
– It fits: solar θ ~ 0.5°, spectrometer θ ~ 1°
– Demonstrate imaging thru Earth’s atmosphere
– It is there everyday
•
Issue: too much radiation → Solar filter
Required
6
Useable Wavebands
PHYSICS
MWIR
1
Atmospheric transmission, Band 2
0.5
0.5
Atmospheric Transmission
Atmospheric transmission, Band 1
1
0
4
4
4.5
5
Wavelength (microns)
4.5
5
5.5
5.5
LWIR
1
1
0.5
0.5
0
0
4 useable wavebands
3.75 – 4.1 µm
4.5 – 4.7 µm
8.2 – 8.5 µm
9.9 – 10.1 µm
0
64
128
8
9
192
256
10
11
Pixel number
0
320
Diminishing Detector Response
7
PHYSICS
Dual-band Spectral Image of
Diameter of Solar Disk
8
PHYSICS
Experimental Setup
Solar image formed by
telescope is allowed to
pass over spectrometer slit
Sun
FPA
Solar Filter
9
PHYSICS
Concatenation of Single Column
Plot of Column 516 (λ = 4.6 µm)
Row #
Frame #
10
PHYSICS
Timing is Important
Plot of Column 600 (λ = 3.95 µm)
Row #
Frame #
11
PHYSICS
Why is the image elliptical?
12
PHYSICS
Data fits to an Ellipse
13
PHYSICS
Circularization Process
•
Concatenate one column
from successive frames
for composite image
•
Find FWHM of each
column
•
•
Find Midpoint of cols
•
Fit top/bottom halves
separately to eq for ellipse
•
•
Find ratio of ellipse axes
Slide each col to align
midpoints
Use ratio to scale
composite image
horizontally
14
PHYSICS
Image can be Circularized
15
Transit Angle and Time
PHYSICS
•
Astronomical calculations
predict:
92 seconds at 90°
131 seconds at 45°
•
Data analysis yields:
132 ± 1 seconds at 42.4° ± 0.5°
•
Gratifying!
16
PHYSICS
Circularized Sun: MWIR
17
PHYSICS
Circularized Sun: LWIR
18
PHYSICS
Median Smoothed Sun: LWIR
Smoothing window of 5 pixels
Smoothing window of 3 pixels
19
PHYSICS
•
•
•
•
•
•
Finding Full Width at Half Max
Must work with bad pixels
Find column mean value
Avg top 10% above
Avg bottom 10% below
Determine halfway
Two methods:
–
–
–
Pixel values
Contiguous pixels
Essentially identical
20
PHYSICS
Solar Diameter vs. λ
122 rows
108 rows
Design value of 15 arcsec for IFOV implies solar diameter of 125 pixels
21
PHYSICS
Sharpness vs. Size
22
PHYSICS
•
Focus Issues
Apparatus is hard to focus on infinity
– Normally take smallest image
– Sun moves
– Therefore solar chord continually growing and
shrinking!
•
Two focus settings used
– First: larger image, but sharper edges
– Second: smaller image, softer edges
• Does magnification change with focus?
23
PHYSICS
Summary: Focus on the Future
• Blackbody at 100 m → done
• Blackbody at 1000 m
– In planning, strobe to facilitate acquisition
– Still not at infinity!
• Star
– Not possible w/o optimal focus
– Recent Dewar modification to facilitate
• Full Moon
– Limited opportunity, once per month
– Tried, but too many clouds
– Plan again for January
• Improved dualband FPA would lead to
dramatic increase in capability, in LWIR!
24