Transcript Midterm_presentation

Multispectral Imager Design
for Nanosatellites
V.H.R.I. Doedee, R. Deerenberg, E.Dokter – Faculties 3mE, AE & EEMCS
8-4-2016
Delft
University of
Technology
Multispectral Imager Design for Nanosatellites
1.
Introduction
Multispectral Imager Design for Nanosatellites
2
Nanosatellite missions so far
• Education
• Technology Demonstration
• No serious remote sensing
Are nano-satellites capable for remote sensing jobs?
Multispectral Imager Design for Nanosatellites
3
Major Constraints
•
•
•
•
•
volume less than 10 cm x 10 cm x 15 cm
power consumption less than 3.0 W
mass less than 1.5 kg
imaging of preferably R, G, B, NIR, MIR, TIR bands
system should survive space environment, including
‘accidental’ sun exposures
• operational life time should be at least five years
Multispectral Imager Design for Nanosatellites
4
How can we achieve this?
• sensors with better quantum efficiencies and lower noise
• deployable instead of rigid optical systems
• in-orbit calibration instead of on-ground
Multispectral Imager Design for Nanosatellites
5
Potential Applications
When using a constellation of nanosatellites
Can increase temporal resolution of remote sensing for:
- Google Earth-like applications
- Precision agriculture
- Climate monitoring
- Disaster prevention & monitoring
- Military intelligence
Multispectral Imager Design for Nanosatellites
6
Contents
•
•
•
•
•
•
Basics of Remote Imaging
CCD & CMOS
Orbit
Noise
Motion Compensation
Questions?
Multispectral Imager Design for Nanosatellites
7
2.
Basics of Remote Imaging
Multispectral Imager Design for Nanosatellites
8
Basics of Remote Imaging
• Sensing in the (optical) EM Spectrum
• Multispectral: multiple bandwiths
•
•
•
•
•
•
UV; ultraviolet, 0.1-0.4 μm
VIS; the visible range, 0.4-0.7 μm
NIR; the near infrared range, 0.7-1.1 μm
SWIR; the short wave infrared range, 1.1-2.5 μm
MWIR; the midwave infrared range, 2.5-7.5 μm
LWIR; the long wave infrared range, 7.5-15 μm
Source: http://www.antonine-education.co.uk/physics_gcse/Unit_1/Topic_5/em_spectrum.jpg
Multispectral Imager Design for Nanosatellites
9
Resolution
• Spatial
• Spectral
• Radiometric
• Temporal
Multispectral Imager Design for Nanosatellites
10
Spatial Resolution
• Smallest measure of seperation between two objects that can
be resolved by the system [T.A. Warner et al, 2009]
• Rayleigh Criterion
d Rayleigh  1.22

DA
H
• Nominal Spatial Resolution
dGIFOV 
Df
f
H   IFOV H
Source: http://www.fas.org/man/dod-101/navy/docs/es310/EO_image/IMG00003.GIF
Multispectral Imager Design for Nanosatellites
11
Minimal Spatial Resolution
Multispectral Imager Design for Nanosatellites
12
Spectral Resolution
• Unitless Ratio:


Source:
http://www.cas.sc.edu/geog/rslab/rscc/mod1/specres.gif
Multispectral Imager Design for Nanosatellites
13
Radiometric Resolution
• How fine a difference in incident spectral radiance can be
measured by the sensor
[T.A. Warner, M. Duane Nellis, G.M. Foody, 2009]
• Quantization of incoming radiation
Multispectral Imager Design for Nanosatellites
14
Temporal Resolution
• Time to refresh images
Multispectral Imager Design for Nanosatellites
15
EM Propagation and Sensors
• Optical sensors act like photon detectors
• Energy needed (bandgap):
E  hf 
hc

• Band gap sets an upper limit
Multispectral Imager Design for Nanosatellites
16
Materials
• Silicon (Si), 0.4-1 μm
• Indium Gallium Arsenide (InGaAs), 0.8-2.8 μm
• Indium Antimonide (InSb), 0.3-5.5 μm
• Mercury Cadmium Telluride (HgCdTe), 0.7-15 μm
Wavelengths absorption of different InGaAs alloys.
Source: http://www.sensorsinc.com/GaAs.html
Multispectral Imager Design for Nanosatellites
17
Beam splitters
• Split light waves in two consecutive beams
• Cube, Plate, Pellicle
• Cube and Plate: only monochromatic light,
heavy
• Pellicle: average transmission 50% (375–
2400 nm), light, little ghosting, sensitive to
vibrations
Source:
http://www.newport.com/store/genproduct.aspx?id=14111
8&lang=1033&Section=Detail
Multispectral Imager Design for Nanosatellites
18
3.
CCD & CMOS
Multispectral Imager Design for Nanosatellites
19
CCD
(Charged-coupled devices)
• ‘Traditional leader’
• Great fill factor (~95%)
• High Quantum efficiency
Multispectral Imager Design for Nanosatellites
20
CMOS
(Complementary metal-oxide-semiconductors)
•
•
•
•
Less circuitry required
Low power consumption
Individually read-out
Cheap
Multispectral Imager Design for Nanosatellites
21
CCD vs CMOS
CCD:
CMOS:
+ create high-quality, low noise images
- more susceptible to noise
+ greater light sensitivity (fill factor, QE)
- lower light sensitivity
- 100 times more power
+ consume little power
- Complex
+ easy to manufacture
- Expensive
+ cheap
Multispectral Imager Design for Nanosatellites
22
4.
Orbit
Multispectral Imager Design for Nanosatellites
23
Orbit
•
Dusk-Dawn Orbit:
No eclipse – No power storage needed.
Same illumination condition surface Earth.
Easy data collection.
•
Altitude = 400 Km:
Typical nanosatellite perigee height.
Lower altitude means better resolution.
Lower altitude also means more drag.
Design lifetime of 5 years achievable.
Multispectral Imager Design for Nanosatellites
24
In-Orbit Calibration
Normally the lens subsystem is calibrated on ground and designed
such that it can withstand the launch without losing focus.
This has major disadvantages
• Loss of focus will always be present.
• Increased risk.
• More mass.
Multispectral Imager Design for Nanosatellites
25
In-Orbit Calibration
Why not perform the calibration in-orbit?
Advantages:
• Less risk.
• Less mass.
• Can use the same mechanism of a possible deployable lens
Disadvantages:
• Less precise calibration
• Not a simple solution!
Multispectral Imager Design for Nanosatellites
26
In-Orbit Calibration
How it works:
The camera makes an image, adjusts the focal length slightly and
takes another image. The two images are then compared. And if
the process is beneficial to the quality of the image then the
process is repeated.
Images are compared either on:
• The satellite itself – More dedicated electronics.
or
• The ground – Issues with communication.
Multispectral Imager Design for Nanosatellites
27
5.
Noise
Multispectral Imager Design for Nanosatellites
28
Signal to Noise Ratio - SNR
The SNR is a measure of the quality of the taken image.
There are two ways to increase the SNR:
• Decrease noise as much as possible
• Increase Integration time to decrease single shot noise such as
Shot Noise
Read out Noise
etc…
Multispectral Imager Design for Nanosatellites
29
Thermal Noise
Thermal Noise is one of the easiest ways to decrease noise.
Materials emit a certain amount of electrons according to their
temperature.
Decreasing the temperature of the sensor would decrease Thermal
Noise significantly. (factor 2 for every 6 degrees of cooling)
Multispectral Imager Design for Nanosatellites
30
Thermal Control
In order to decrease Thermal Noise, the sensor must be cooled.
But:
Thermal control of a nanosatellite is difficult due to it’s limited size
• No Active control can be applied
• Only Passive control is an option:
Thus the satellite must be coated with a coating with high
emissivity and a low absorption factor.
Multispectral Imager Design for Nanosatellites
31
Further decreasing SNR
Other types of noise are dependent on the sensor and electronics.
Such as:
• Quantum Efficiency (QE) – Measure of efficiency of the sensor,
this value should be as high as possible within the required
spectrum.
• Amplifying noise
• 1/f noise
• Etc…
This should be as low as possible but this will increase the cost of
the S/C.
Multispectral Imager Design for Nanosatellites
32
Integration Time
Some values of noise are a single event values, these do not
increase with measurement time.
When the time in which we view an object (Integration Time) is
increased, the SNR goes up.
This is however not a simple task since the S/C is moving with
respect to the ground: Blurring effect.
Multispectral Imager Design for Nanosatellites
33
Blurring Effect?
Multispectral Imager Design for Nanosatellites
34
6.
Motion Compensation
Multispectral Imager Design for Nanosatellites
35
Motion Compensation
Two ways in doing so:
• Mechanical movement of the lens subsystem to track a point on
the ground.
• Time Delay Integration - TDI.
Multispectral Imager Design for Nanosatellites
36
Mechanical movement method.
In order to view the same point on the ground a tilting mirror can
be placed in front of the lens subsystem to reflect the rays of
light.
The motion of this mirror has to be synchronized with the
movement of the S/C w.r.t the ground.
•
•
•
•
This increases complexity since moving parts are necessary
Increase in power consumption
Increase in development cost
But large integration time is achievable.
Multispectral Imager Design for Nanosatellites
37
Time Delay Integration - TDI
Time Delay Integration is a method which uses no moving parts.
Instead it uses an array of pixels.
When the satellite passes a point on the ground, the first pixel
takes a measurement, and the pixel is read out.
Due to the motion of the spacecraft, the same point can be seen
by the next pixel in the direction of flight after a small time step.
This holds for the entire pixel array.
This method is also called push broom scanning
Multispectral Imager Design for Nanosatellites
38
Time Delay Integration - TDI
Some advantages/disadvantages:
• Smaller Integration time achievable
• No CCD sensor can be used, only CMOS
• Much lower mass
• Tested and proven method.
Multispectral Imager Design for Nanosatellites
39
Questions?
Multispectral Imager Design for Nanosatellites
40