© The Aerospace Corporation 2009

PICASSO –

An End-to-End Image Simulation Tool

for Space and Airborne Imaging Systems

02 Apr. 2009

Steve A. Cota, Jabin T. Bell, Richard H. Boucher,

Tracy E. Dutton, Chris J. Florio, Geoffrey A. Franz,

Thomas J. Grycewicz, Linda S. Kalman, Robert A. Keller, Terence S.

Lomheim, Diane B. Paulson, and Timothy S. Wilkinson

2

What is PICASSO?

Parameterized Image Chain Analysis and Simulation SOftware– One of a class of End-to-End (ETE) image simulation (aka, Image Chain Analysis) tools.

– Development began in 1999.

– Among goals were GUI & architecture facilitating code expansion & maintenance.

PICASSO predicts system-level performance of electro-optical imaging systems– Not a detailed optics, detector, etc. model, but a system-level roll-up of component sub-

system performance metrics.

– Requires high quality “truth” image & sensor design description as inputs.

– Outputs are simulated imagery & figures of merit (e.g., SNR, MTF).

PICASSO may be used throughout the life of a satellite program.– Most often used early in program, when design parameters are still in trade.

– Often used to gain insight into contractor modeling techniques.

– May be applied later in program to predict system-level performance of as-built hardware.

3

GENSAT Design Parameters (1/2)

500Altitude [km]

0.5Nadir resolution [m]

0.5Ground sample distance (GSD) at nadir [m]

0.95Filter transmission (average over panchromatic band)

0.80Optical throughput (average over panchromatic band)

0.10Wavefront error correlation length [fraction of primary aperture]

0.10Wavefront error [r.m.s. waves @0.6328 μμμμm]1.0Optical Q at 0.65 μμμμm10Effective focal length (EFL) [m]

0.10Secondary aperture (obscuration) diameter [m]

0.65Primary aperture diameter [m]

CassegrainOptical design

Panchromatic (0.45 – 0.80)Optical bandpass [μμμμm]

ValueParameter

PICASSO Example: GENSAT

– Generic high resolution commerical imaging satellite

– Representative of the current 0.5m ground resolution class of systems.

– Representative of a “Conceptual Design” level of maturity.

4

GENSAT Design Parameters (2/2)

16, 32, 48, 64, 80, 96TDI modes

36, 18, 9, 4.5Line rate modes [KHz]

0.5Wiener filter gain

1.5A/D maximum voltage [V]

11Analog-to-Digital Convertor [bits]

0.65Quantum efficiency (average over panchromatic band)

1500Pixel full well depth parameter [electrons/μμμμm2]

50Total detector noise (dark, readout, etc.) [electrons]

10CCD detector pitch [μμμμm]Scanning 4-phase CCDFocal plane array

ValueParameter

5

Typical PICASSO Imaging Chain

High

Resolution

Input Scene

in Unknown

Units

Apply System Transfer

Function (STF)

Resample to

Sensor GSD

Convert from Radiance to

Photoelectrons

Add Detector &

Photon Noise

Analog-to-Digital

Conversion

Sharpening via

Wiener Filter

Output Scene

& Associated

Metrics

PICASSO

Rescale from Reflectance to

Inband Top of Atmosphere

Radiance at Desired Sun &

Sensor Elevation Angles via

MODTRAN4v3r1

Convert Input

Scene to

Reflectance

Convert from Electrons

to Voltage

Setup

Viewing

Geometry

6

GENSAT Simulation Chain

High

Resolution

Input Scene

in Unknown

Units

Apply System Transfer

Function (STF)

Resample to

Sensor GSD

Convert from Radiance to

Photoelectrons

Add Detector &

Photon Noise

Analog-to-Digital

Conversion

Sharpening via

Wiener Filter

Output Scene

& Associated

Metrics

PICASSO

Rescale from Reflectance to

Inband Top of Atmosphere

Radiance at Desired Sun &

Sensor Elevation Angles via

MODTRAN4v3r1

Convert Input

Scene to

Reflectance

Convert from Electrons

to Voltage

Setup

Viewing

Geometry

7

Rescaling to Top of Atmosphere Radiance (1/2)

MODTRAN4 reflectance dependence is fit

using the FLAASH[1] model:

( )( )( )( )

( )( )( ) λ

λ

λ

λ

λλ

ρ

ρ

ρ

ρC

Sx

xB

Sx

xAxL

b

b

b

+−

+−

=11

[1] Adler-Golden et al., “FLAASH, A MODTRAN4 Atmospheric Correction Package for Hyperspectral Data

Retrievals and Simulations”, Summaries of the Seventh JPL Airborne Earth Science Workshop, JPL Publication

97-21, 1, pp. 9-14 (1998).

0.0E+00

5.0E-03

1.0E-02

1.5E-02

2.0E-02

2.5E-02

3.0E-02

3.5E-02

4.0E-02

0.40 0.50 0.60 0.70 0.80 0.90

Wavelength [m]

W/cm

2/sr/ m

A

B

C

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.40 0.50 0.60 0.70 0.80 0.90

Wavelength [m]

Spherical Albedo - S

Example fit coefficients for overhead sun,

90o sensor elevation angle, mid-latitude

summer atmosphere, 23 km visibility rural

aerosol model.

8

GENSAT Simulation Chain

High

Resolution

Input Scene

in Unknown

Units

Apply System Transfer

Function (STF)

Resample to

Sensor GSD

Convert from Radiance to

Photoelectrons

Add Detector &

Photon Noise

Analog-to-Digital

Conversion

Sharpening via

Wiener Filter

Output Scene

& Associated

Metrics

PICASSO

Rescale from Reflectance to

Inband Top of Atmosphere

Radiance at Desired Sun &

Sensor Elevation Angles via

MODTRAN4v3r1

Convert Input

Scene to

Reflectance

Convert from Electrons

to Voltage

Setup

Viewing

Geometry

9

GENSAT Modulation Transfer Function (MTF)

In-scan component & system polychromatic MTFs for excess smear = 0., nadir viewing, line rate 18 kHz/TDI by 32 mode

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

0.00 0.50 1.00 1.50 2.00 2.50 3.00

Spatial Frequency / Nyquist Frequency

Modulation Transfer Function (MTF)

Detector

Optics

Smear

CCD Diffusion

Wavefront Error

System Transfer Function

10

GENSAT Simulation Chain

High

Resolution

Input Scene

in Unknown

Units

Apply System Transfer

Function (STF)

Resample to

Sensor GSD

Convert from Radiance to

Photoelectrons

Add Detector &

Photon Noise

Analog-to-Digital

Conversion

Sharpening via

Wiener Filter

Output Scene

& Associated

Metrics

PICASSO

Rescale from Reflectance to

Inband Top of Atmosphere

Radiance at Desired Sun &

Sensor Elevation Angles via

MODTRAN4v3r1

Convert Input

Scene to

Reflectance

Convert from Electrons

to Voltage

Setup

Viewing

Geometry

11

Example PICASSO Study: Effects of Wavefront Error (WFE) (1/2)

Sharpening via Wiener Filter – Poorer SNR for WFE case at high spatial

frequencies leads to less complete restoration

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

0.00 0.20 0.40 0.60 0.80 1.00

Spatial Frequency / Nyquist Frequency

Modulation Transfer Function (MTF)

WFE = 0.00 waves rms

WFE = 0.10 waves rms

0.90

0.91

0.92

0.93

0.94

0.95

0.96

0.97

0.98

0.99

1.00

0.00 0.20 0.40 0.60 0.80 1.00

Spatial Frequency / Nyquist Frequency

Modulation Transfer Function (MTF)

WFE = 0.00 waves rms

WFE = 0.10 waves rms

12

4.70

4.72

4.74

4.76

4.78

4.80

4.82

4.84

4.86

4.88

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90

Solar Elevation Angle [deg]

GIQE 4.0 NIIRS

0.00 pix/msec/TDI

0.10 pix/msec/TDI

0.20 pix/msec/TDI

Example PICASSO Study: TDI Mode Switchpoints

PICASSO & General Image Quality Equation (GIQE) 4.0[1]

used to derive TDI mode switchpoints as solar illumination declines

[1] Leachtenauer, J. C., et al., ‘‘General Image Quality Equation: GIQE,’’ Appl. Opt. 36, 8322–8328 (1997).

324864

3248

64

80TDI x 96