Share Email Print

Spie Press Book

Electro-Optical Imaging System Performance, Sixth Edition
Author(s): Gerald C. Holst
Format Member Price Non-Member Price

Book Description

This sixth edition emphasizes staring array analysis and provides:

• NVIPM, TRM4, and TOD comparisons
• Frequency (MTF) versus spatial (pixels on target) analysis
• Two-dimensional versus two-directional analyses
Fλ/d approach to modeling and system resolution
• In-band and out-of-band sampling artifacts
• Numerous trade studies

Copublished with JCD Publishing

Book Details

Date Published: 10 April 2017
Pages: 407
ISBN: 9781510611023
Volume: PM278

Table of Contents
SHOW Table of Contents | HIDE Table of Contents

Table of Contents

Symbols and Acronyms


1 Introduction
1.1 System modeling
1.2 System applications
1.3 Nomenclature
1.4 System modeling
1.5 Staring array modeling
1.6 References

2 Imaging System Design
2.1 IR detectors
      2.1.1 Detector classification
      2.1.2 Specific detectors
2.2 Scanners
2.3 Staring arrays
2.4 IR system evolution
      2.4.1 Common module systems
      2.4.2 EOMUX systems
      2.4.3 EMUX systems
      2.4.4 Second-generation systems
      2.4.5 Staring array systems
      2.4.6 Third-generation systems
2.5 Visible system detectors
      2.5.1 CCDs
      2.5.2 CMOS
2.6 Image intensifiers (I2 devices)
2.7 Intensified solid-state cameras
2.8 High-definition television
2.9 References

3 Fill Factor and Field of View
3.1 Pixels, datels, and disels
3.2 Fill factor
3.3 Field of view
3.4 Optical blur


4 Linear System Theory
4.1 Linear filter theory
4.2 Optical transfer function
4.3 MTF definition
4.4 Spatial frequency
      4.4.1 Object and image spatial frequencies
      4.4.2 Observer spatial frequency
4.5 Separability
4.6 Phase shifts
4.7 Final word
4.8 References

5 Staring Array MTFs
5.1 Optics OTF
      5.1.1 Diffraction-limited OTF
      5.1.2 Central obscuration
      5.1.3 Rectangular aperture
      5.1.4 Non-circular apertures
      5.1.5 Anamorphic optics
      5.1.6 Aberrations
      5.1.7 Defocused optics
      5.1.8 Tolerancing
5.2 Detectors
      5.2.1 Rectangular detectors
      5.2.2 Non-square detectors
      5.2.3 Visible color cameras
      5.2.4 Diffusion
      5.2.5 Charge transfer efficiency
      5.2.6 TDI mismatch
5.3 Flat-panel displays
5.4 Eye MTF
5.5 References

6 Environmental MTFs
6.1 Motion
      6.1.1 Linear motion
      6.1.2 Sinusoidal motion
      6.1.3 Random motion (jitter)
      6.1.4 Low-frequency motion
6.2 Atmospheric turbulence
      6.2.1 CN2
      6.2.2 Fried's coherence diameter
      6.2.3 Horizontal path length
      6.2.4 Slant path
      6.2.5 Turbulence MTF
6.3 Aerosol MT
6.4 References

7 Two-Dimensional MTF
7.1 Optics
7.2 Motion
      7.2.1 Linear motion
      7.2.2 Sinusoidal motion
      7.2.3 Random motion
7.3 Detector
7.4 Digital processing
7.5 Displays
7.6 The eye
7.7 Atmospheric turbulence
7.8 References

8 Sampling
8.1 Sampling theory
8.2 Staring array sampling frequency
8.3 Reconstruction
      8.3.1 Spurious response
      8.3.2 MTF "squeeze"
      8.3.3 Resolved cycle contraction
8.4 Microscan
8.5 Super-resolution reconstruction
8.6 Sample-scene phase
8.7 Color filter arrays
8.8 Sampling artifacts
8.9 Can you see the sampling lattice?
8.10 Digital data representation
8.11 References

9 Image Processing
9.1 z-transform
9.2 Digital filters
      9.2.1 Unsharp filter
      9.2.2 Averaging filter
9.3 Interpolation (electronic zoom)
      9.3.1 Ideal interpolator
      9.3.2 Lanczos interpolator
      9.3.3 Pixel replication
      9.3.4 Linear interpolation
      9.3.5 Bilinear interpolation
9.4 Line-to-line interpolation
9.5 A noise reduction algorithm
9.6 Image restoration
9.7 Image enhancement
9.8 Final word
9.9 References


10 Detector Responsitivity
10.1 Filter transmission
10.2 Classical semiconductors
10.3 Novel semiconductors
10.4 Thermal detectors
10.5 Specific detectivity
10.6 References

11 Radiometry
11.1 Radiative transfer
11.2 Planck's blackbody law
11.3 Camera formula
      11.3.1 Infrared sources
      11.3.2 Visible, NIR, and SWIR sources
11.4 Point source
11.5 Photometry
11.6 Normalization
11.7 References

12 Signal-to-Noise Ratio
12.1 Common noise sources
      12.1.1 Shot noise
      12.1.2 Dark current
      12.1.3 Multiplexer noise
      12.1.4 Quantization noise
      12.1.5 Non-scene photons
      12.1.6 Spatial noise
12.2 NEDT
      12.2.1 Background temperature
      12.2.2 FPN
      12.2.3 Variable integration time
      12.2.4 Warm optics
      12.2.5 Uncooled systems
12.3 NEΔρ
12.4 NEI
12.5 SNR optimization
12.6 Sun glints
12.7 Signal-to-clutter ratio
12.8 Three-dimensional noise model
12.9 Noise figure
12.10 Real systems
12.11 References


13 Atmospheric Effects
13.1 Atmospheric constituents
      13.1.1 Water vapor
      13.1.2 Aerosols
13.2 Atmospheric codes
      13.2.1 LOWTRAN, MODTRAN, and HITRAN
      13.2.2 MATISSE
13.3 Visible average transmittance
13.4 Infrared average transmittance
      13.4.1 MWIR versus LWIR
      13.4.2 Navy model
      13.4.3 Land-based systems: horizontal path
      13.4.4 Land-based systems: slant path
13.5 Weather conditions
      13.5.1 Rain and snow
      13.5.2 Probability of occurrence
13.6 Scattering and path radiance
      13.6.1 Infrared
      13.6.2 Visible contrast transmittance
13.7 Obscurants
13.8 References

14 Target Signatures
14.1 Target contrast
14.2 Targets in the IR
      14.2.1 Area-weighted ΔT
      14.2.2 Diurnal variations
      14.2.3 Environmental modifiers
      14.2.4 Active targets
14.3 Targets in the visible/NIR/SWIR
14.4 Target signature modeling
14.5 References


15 Resolution
15.1 Analog metrics
      15.1.1 Optical resolution
      15.1.2 NIIRS
15.2 Sampled data systems
      15.2.1 Spot size ratio
      15.2.2 Ensquared power
15.3 Schade's equivalent resolution
15.4 Fλ/d
15.5 References

16 Image Quality
16.1 Mathematical metrics
16.2 MTF
16.3 Perceived resolution
16.4 MTFA
16.5 Subjective quality factor
16.6 Square-root integral
16.7 Targeting task performance
16.8 Resolution versus perceivable detail
16.9 References

17 System Performance Models
17.1 NVL 1975 model
17.2 FLIR92
      17.2.1 Frame integration
      17.2.2 Viewing distance
      17.2.3 Two-dimensional MRT
17.3 NVTherm
      17.3.1 Eye contrast threshold function
      17.3.2 Perceived SNR
17.4 NVThermIP
      17.4.1 Gain and level
      17.4.2 Eye integration time
      17.4.3 Predicted MRT
17.5 NVIPM
      17.5.1 Legacy modeling
      17.5.2 Eye integration time
      17.5.3 ΔT versus TB
17.6 Moderate aspect ratio targets
17.7 TRM
17.8 Triangle orientation discrimination
17.10 Model comparisons
17.11 References

18 Additional Metrics
18.1 Pixels are not cycles
18.2 Hot spot detection
18.3 Pixels on target
18.4 Search
18.5 General image quality equation
18.6 References


19 Target Discrimination
19.1 Discrimination definitions
19.2 Cycles on target
19.3 "Johnson" two-dimensional criteria
19.4 NVTherm/NVThermIP
      19.4.1 Conversion of N50 to V50
      19.4.2 V50 values
19.5 NVIPM
      19.5.1 Target sets
      19.5.2 V50 values
19.6 Characteristic dimension
19.7 Alphanumeric readability
19.8 References

20 Range Predictions
20.1 Target transfer performance function
20.3 1975 model/FLIR92/NVTherm
20.4 TRM
20.5 NVThermIP/NVIPM
20.6 Clutter
20.7 Field demonstrations
20.8 References

21 Trade Studies
20.1 1975 model, FLIR92, and NVTherm
21.2 NVThermIP approximation
21.3 NVIPM equations
      21.3.1 Gain
      21.3.2 Fλ/d
      Effect of noise on acquisition range
21.4 Environmental MTFs
      21.4.1 Atmospheric turbulence
      21.4.2 Line-of-sight stabilization
21.5 Observer viewing distance
21.6 Two fields of view
21.7 Gradient analysis
21.8 References

      Focal Ratio


Preface to the Sixth Edition

Microsoft Word was my real frustration in producing this edition. The prior (5th) edition was written with Office 2003. Office 2003 equations are not compatible with Office 2013 — so each equation (all 442 of them) had to be retyped. The aggravation does not end there. No keyboard character can be on the same line as the equation. So, the equations are placed in a 2×1 table with the equation in the center cell and the equation number in the righthand cell. Now comes the editing. Text, tables, references, and figures have been rearranged for efficient layout.

Since this text is about systems, it is sometimes difficult to decide in which chapter a specific topic belongs, since everything is intertwined. For example, integration time affects linear motion (Chapter 6), the number of photoelectrons (Chapter 11), and the system noise (Chapter 12). As the book evolved the chapter sequence changed numerous times, ending up with six sections: 1) The basics, 2) Imaging system chain analysis: MTF approach, 3) Signal to noise ratio, 4) Targets and backgrounds, 5) Image quality metrics, and 6) Acquisition range. Words are important: can and will have very different meanings. Likewise, suggests means the statement might be true whereas is means the truth.

This book focuses on staring arrays. Readers can find scanning array equations in the (out-of-print) 5th edition. Up-to-date references (latest published September 2016) have been added. A few recent ones are somewhat dubious. Time will tell if these papers are fact or fiction. But, that is the nature of research. What appears to be a good idea today is trashed by future research. You can go to my earlier editions and easily say, "That is not true!" Missing is the operative word today: That is not true TODAY.

The first edition appeared in 1995 and has been updated over the years, with over 5,700 copies sold. Hard to believe that 5,700 folks are interested in this niche area. So, I say to my friends: "During an archeological dig one thousand years from now, this book appears and the scientists will exclaim 'What was he thinking?'" Which is, exactly, the topic of this book.

OK — What is new in the sixth edition?

The U.S. Army models have evolved over the years to keep abreast with hardware technology changes, laboratory data, and field performance. There are four primary U.S. models: 1) NVThermIP predicts the performance of systems operating in the MWIR and LWIR thermal bands; 2) IINVD is used for image intensifier direct-view goggles; 3) SSCamIP models reflective band cameras; and 4) IICam addresses indirect view I2 sensors where a tube is coupled with a detector array. The four models have been combined into NVIPM (Night vision integrated performance model), which was released in May 2013.

Targets in the visible, NIR, SWIR, MWIR, and LWIR spectral bands have different spectral components and nomenclature. Thus, NVIPM contains a module for each spectral band. The detected signal and subsequently displayed image do not contain any sensor spectral information and the post-detector signal is common to all imaging systems. Since NVIPM includes visible and SWIR systems, a few sections for CCD, CMOS, and SWIR sensors have been added.

Prior to NVIPM, each variable change required re-running the model. NVIPM provides an extensive trade study capability that allows users to compare system performance by varying any combination of input, output, and component values. With NVIPM, the variables are run in batch mode and the output (e.g., range) is plotted as a function of any input variable. Also included is a gradient feature that indicates which input variables affect the output, in descending order. The gradient feature is extremely useful for subsystem tolerancing.

Probably the most valuable chapter is Chapter 21, Trade studies. The various variables can be considered as slices through a multi-dimensional space. Since it is not possible to illustrate more than three dimensions at a time, each tradeoff analysis represents only one plane through this space. Each provides a different view of overall optimization. For example, if a system is detector-limited, modifying the optics will have minimal effect on acquisition range. If turbulence limited or motion limited, no system change will appreciably affect performance.

Because the U.S. models have continually changed, the Europeans are developing the European computer model for optronic system performance prediction (ECOMOS) based on the German thermal range model (TRM) and the Dutch triangle orientation discrimination (TOD). Each model (NVIPM, TRM, TOD) provides a different target acquisition range. At this juncture, it is unknown which is "correct" although the respective authors think they know.

It is hard work writing a technical book; taking somewhere between 6 months and a year glued to a computer. My wife had said many times, "I want you back." Marilyn, I am back!

Orges Furxhi, St. Johns Optical, and Chris Dobbins, AMRDEC, provided valuable technical reviews. Doug Marks, itinerant editor, provided many of the drawings and final editing. I appreciated Doug's "instant" response to my requests for drawing modifications.

Gerald C. Holst
March 2017

© SPIE. Terms of Use
Back to Top