Review of Night Vision Technology

OPTOELECTRONICS REVIEW 21(2), 153181
DOI: 10.2478/s1177201300893
Review of night vision technology

INVITED PAPER
K. CHRZANOWSKI*1,2
1Institute
of Optoelectronics, Military University of Technology, 2 Kaliskiego Str., 00908 Warsaw, Poland

2Inframet, 24 Graniczna Str., Kwirynw, 05082 Stare Babice, Poland
Night vision based on technology of image intensifier tubes is the oldest electrooptical surveillance technology. However, it
receives much less attention from international scientific community than thermal imagers or visible/NIR imagers due to se
ries of reasons. This paper presents a review of a modern night vision technology and can help readers to understand sophis
ticated situation on the international night vision market.
Keywords: night vision technology, image intensifier tubes.
List of abbreviations
1. Introduction
ANVIS
Humans achieve ability to see at night conditions by using

several different imaging systems: night vision devices
(image intensifier systems), thermal imagers, SWIR ima
gers, and some more sensitive visible/NIR (CCD/CMOS/
ICCD/EMCCD) cameras. However, due to historical rea
sons, night vision technology is usually understood as night
vision devices.
Night vision devices (NVDs) are apparently simple sys
tems built from three main blocks: optical objective, image
intensifier tube, and optical ocular (Fig. 1).
The task of the optical objective is to create low intensity,
invisible image of the observed scenery at input plane of the
image intensifier tube. The latter tube consisting of a pho
tocathode, an anode in form of a phosphor screen, and other
components, intensifies an input lowluminance image into
a brighter image created on the anode (screen). The latter im
age is seen by human observer using the optical ocular.
Design of NVDs is apparently easy because crucial
modules like image intensifier tube, optical objectives, opti
cal oculars (eyepieces) are available on the market from
dozen or more sources. However, in spite of this apparent
design simplicity, the process of creating output image by
these imaging systems is quite sophisticated. Many design
rules must be well understood by manufacturers to deliver
aviator night vision imaging system (a term used

commonly for binocular night vision goggles),
CCD
charge couple device (a technology for constructing
integrated circuits that use a movement of electrical
charge by shifting signals between stages within
the device one at a time),
CCTV
close circuit television (type of visible/NIR cameras
used for short range surveillance)
CMOS
complementary metaloxidesemiconductor (a tech
nology that uses pairs of ptype and ntype metal
oxide semiconductor field effect transistors for con
structing image sensors
CRT
cathode ray tube (a vacuum tube containing an electron
gun and a phosphor screen used to generate images)
EMCCD electron multiplying charge coupled device
fc
foot candela
fL
foot lambert
ENVG
enhanced night vision goggles
EBAPS electron bombarded active pixel sensor
FOM
figure of merit
FOV
field of view
HUD
headup display
ICCD
intensified CCD (a technology that uses imaging mo
dules achieved by combing image intensifier tube
with CCD sensor)
IIT
image intensifier tube
lp/mm
line pair per millimeter
lp/mrad line per miliradian
MCP
micro channel plate
MIL standard a United States defence standard, often called
a military standard
NIR
near infrared
NVD
night vision device
NVG
night vision goggles
RMS
root mean square
SNR
signal to noise ratio
SWIR
short wave infrared
TFT LCD thin film transistor liquid crystal display.
*email:
Fig. 1. Block diagram of night vision monocular.
[email protected]
OptoElectron. Rev., 21, no. 2, 2013
K. Chrzanowski
153

high performance NVDs. Every manufacturer of NVDs
must carry out some kind of performance/cost optimization
that requires deep knowledge of process of influence of dif
ferent modules on quality of final image and functionality of
final night vision device.
Night vision devices have a relatively long history in
comparison to thermal imagers or visible/NIR cameras.
First NVDs have been developed before the end of the Sec
ond World War [1] in situation when the first thermal
imagers have been developed at the beginning of 1970s; and
the first visible/NIR cameras based on modern solid state
technology appeared on market in 1980s. However, NVDs
have received much less attention from international scien
tific community than thermal imagers or visible/NIR cam
eras due to two reasons. First, technology of image intensi
fier tubes (crucial module of NVDs) has been developed
mostly by big manufacturers, not by scientific institutes.
The manufacturers have a natural unwillingness to free
access publications in contrast to scientific institutes. Sec
ond, it was predicted many times that night vision technol
ogy will demise in near future due to competition from more
modern surveillance imagers like thermal imagers, low light
visible/NIR cameras (ICCD/EMCCD cameras), or more re
cently SWIR imagers. Therefore, night vision technology
has been treated by most scientists as rather old, unfashiona
ble technology in the last several decades.
This low interest of scientific community in night vision
technology resulted in a small number of specialist litera
ture, a certain chaos in available literature, conflicting
claims of superiority of different types of image intensifier
tubes, marketing arguments promoted by different manu
facturers and repeated by some scientists. Even specialists
in night vision technology can have real problems with
objective evaluation of modern night vision devices on the
basis of available literature data.
This paper presents a review of modern night vision
technology with aim to present a logical, consistent vision
of this technology divided using a series of criteria. NVDs
are classified in dependence on design configuration of
NVD, type of image intensifier tube, type of night vision
optics, targeted market, type of aviator NVD, and compati
bility to aviation regulations. Short review of parameters
and development trends of night vision technology are
presented, as well.
Over 99% of NVDs offered on world market can be

treated as equivalents of the models listed above and shown
in Figs. 25.
Fig. 2. Night vision goggles: (a) binocular night vision goggles type
AN/AVS9 (ITT Night Vision [4]), and (b) monocular night vision
goggles type BIG 25 (Vectronix AG [5]).
2. Design configuration
There is no internationally accepted division of NVDs. The
same types of NVDs can have different names in different
literature sources. Here we will follow a division and a ter
minology used at websites of two big manufacturers of
NVDs and divide modern night vision devices into four
basic types [2,3]:
1. Night vision goggles;
2. Night vision monoculars;
3. Night vision sights;
4. Night vision binoculars.
154
Fig. 3. Night vision monocular type Tarsius (Vectronix AG [5]).
2013 SEP, Warsaw
Fig. 4. Night vision sight type Trident (ATN Corp. [2]).
Fig. 5. Night vision binoculars: (a) twochannel ATN Night Ra

ven2 binoculars (ATN Corp. [2]), and (b) singlechannel Diana 3x
binoculars (Optix Corp. [6]).
The first two groups of NVDs (goggles, monoculars) are

basically devices of a wide field of view (FOV) similar to
human vision (FOV about 40, magnification equal to one).
These NVDs can be treated as human eyes of improved sen
sitivity.
Binocular night vision goggles enable observation using
two eyes to achieve stereo vision (threedimensional (3D)
vision). In other words, human using binocular night vision
goggles can achieve perception of depth from two slightly
different projections of the world onto the retinas of the two
eyes. Binocular night vision goggles are typically used by
pilots, drivers or other people who need a 3Dvision of sur
rounding scenery at night conditions.
Monocular night vision goggles can be treated as a chea
per version of the earlier discussed binocular night vision
goggles. Two costly image intensifier tubes are replaced by
one tube. A comfortable twoeye observation is still possi
ble. Some depth perception is still achieved even during
a singlechannel observation.
In case of night vision monoculars the simplification

process goes even further. The monoculars are practically
onechannel binocular night vision goggles. Price is reduced
by factor at least two in comparison to binocular night
vision goggles. Additional advantage is small size and mass
of these devices.
The last two groups of NVDs (sights, binoculars) are
basically devices of narrow field of view (FOV from about
4 to about 13, magnification from about 3 to about 10).
These devices can be treated as human eyes of improved
sensitivity equipped with magnifying optical scope.
Night vision sights (called also often night vision
scopes) are generally monoculars of a narrow FOV that pro
vide magnification of an image perceived by a human oper
ator by a factor from 3 to 10. These devices are typically
attachable to weapons.
Night vision binoculars are night vision goggles built by
using bigger optical objectives of a longer focal length. The
binoculars enable magnification of an image perceived by
a human operator by a factor from 3 to 10 like typical day
level binoculars. If built using two separate optical channels
then the binoculars offer also stereoscopic vision.
So far, we listed four different basic types of NVDs that
look and work differently from final user point of view.
However, night vision devices can be divided in different
way taking into account their design. From the latter point
of view differences between goggles and binoculars, and
between monoculars and sights are small. The differences
are caused only by one module; optical objective. Because
of this minor technical difference, the borders between ear
lier discussed groups of NVDs are fluid. Night vision gog
gles can be easily converted to night vision binoculars if
optical objectives are exchanged for bigger objectives of
a longer focal length or some afocal adapters are added.
Then, night vision monoculars can be converted into night
vision sights by exchanging objective and by adding some
mechanics that make possible to attach the monocular to
weapons. There are many such NVDs on the market.
Therefore, from designer point of view, NVDs are
divided into three basic types (Fig. 6):
1. Binochannel NVDs;
2. Mixed channel NVD;
3. Monochannel NVD.
Technical differences are significant between them and
it is not possible to convert easily one type of NVDs to
another types of NVDs. The latter division is based on more
logical ground and presents deeper differences than the ear
lier discussed commercial division. However, the first divi
sion is more popular and should be remembered if someone
wants to understand terminology used by manufacturers of
NVDs.
So far NVDs have been divided using easily visually
noticeable differences (number of objectives, number of
oculars, size of objective, field of view). Now, let us discus
more subtle, but also more important differences between
different NVDs.
K. Chrzanowski
155
Fig. 6. Block diagrams of three types of night vision devices: (a) bino channel NVDs, (b) mixedchannel NVDs, and (c) mono channel NVDs.
3. Image intensifier tubes

Image intensifier tubes (IITs) are vacuum tubes that amplify
a low lightlevel image to observable levels. The incoming
light is converted into photoelectrons through a photoca
thode of the tube. Next, highly intensified photoelectrons
strike the phosphor screen (anode) and a bright image is cre
ated that human can easily see.
Image intensifier tubes are the most important compo
nent of night vision devices and typical classification of
night vision devices is based on a tube classification.
IITs are typically divided into several generations in
dependence of the method to amplify incoming light (pho
tocathode material, tube design structure) as the basic crite
rion. IITs can be also classified by using other criteria like
type of input optics, type of output optics, phosphor screen,
photocathode size, or tube performance.
Five parameters of IITs must be defined to enable preci
sion discussion about division of these tubes:
1. Radiant sensitivity is a ratio of current induced into
a photocathode (in mA units) of tested tubes by incom
ing light (in Watt units) for a specified wavelength.
2. Luminous sensitivity is a ratio of current induced into
a photocathode (in A units) of the tested tube by flux
(in lumen units) of incoming polychromatic light of col
our temperature equal to 2856 K.
3. Luminance gain is a ratio of luminance of output image
(tube screen) to luminance of input image (tube photo
cathode). Measurement is done using light source of co
lour temperature equal to 2856 K. Luminance gain can
be presented in several ways: in cd/lxm2 units (candela
per lux times square meter), in lm/lm units (lumen per
lumen), or in fL/fc (footlambert per foot candela). Im
age generated by a tube of low luminance gain looks
darker than image generated by a tube of high luminance
gain at the same input illuminance conditions.
4. Resolution is defined as a spatial frequency of a minimal
3bar pattern of USAF 1951 target that can be resolved
by an observer. Resolution is presented in lp/mm (line
pairs per millimeter) units. Simulated images of USAF
156
1951 target of two tubes of different resolution at input

illumination about 3 mlx generated using a Nightmet
computer simulator program are shown in Fig. 7 [7].
Nowadays, resolution of typical IITs available on mar
ket is about 5057 lp/mm; resolution of the best tubes
can reach level of 81 lp/mm.
5. Signal to noise ratio (SNR) is a ratio of two components
of a light signal emitted by a small part of a tube screen:
average signal to root mean square signal (noise). The
output signal is generated by illuminating a small part of
photocathode (diameter 0.2 mm) at typical level of 108
lx. Simulated images of USAF 1951 target of two tubes
of different SNR at input illumination about 0.3 mlx ob
tained using Nightmet computer simulator program [7]
are shown in Fig. 8. Nowadays, SNR of typical IITs
available on market is about 1822; SNR of the best
tubes can reach level of 30.
3.1. Generations of image intensifier tubes

First night vision devices were developed during the Second
World War [1]. The technology of image intensifier tubes
has progressed very significantly since that time. This prog
ress can be described using different divisions but the most
Fig. 7. Magnified image of USAF 1951 resolution target generated

by two tubes of different resolution: (a) tube of resolution 64 lp/mm
and (b) tube of resolution 40.3 lp/mm (case of high input illumina
tion).
2013 SEP, Warsaw
Fig. 8. Magnified image of USAF 195,1 resolution target generated

by two tubes of different SNR: (a) tube of SNR equal to 28 and (b)
tube of SNR equal to 18 (case of low input illumination).
popular is division onto generations. US military has dic

tated the name of the generation of IITs over the last five
decades. It should be however remembered that the division
on generations is assumed in USA and presents US point of
view on the night vision technology.
There are so far four enerations of image intensifier tubes:
Gen 0, Gen 1, Gen 2, Gen 3 at least according to the official
US terminology. In general, generation numbering is related
to significant changes in design of IITs that improve (with
some exceptions) performance of these tubes.
As we see in Fig. 9, different generations of IITs use dif
ferent photocathodes. There is a big positive difference bet
ween S1 photocathode used by Gen 0 tubes and S10
photocathode used by Gen 1 tubes. The positive difference
between S25 photocathode used by Gen 2 tubes and S10
photocathode used by Gen 1 tubes is not so obvious. How
ever, we must take into account that critical parameter of
photocathodes of IITs luminous sensitivity is measured
using light sources of 2856K colour temperature. This mea
surement method (simulating to some degree real applica
tions) strongly favours photocathodes more sensitive in near
infrared range in this case S25 photocathode. Finally,
GaAs photocathodes used in Gen 3 tubes are again clearly
Fig. 9. Spectral sensitivity of typical photocathodes: S1(Gen 0),

S20 (Gen 1), S25 (Gen 2), GaAs (Gen 3) and relative Planck
curve.
more sensitive than S25 photocathodes used in tubes of

previous generation.
It should be noted that radiant sensitivity of photocatho
des type S1, S20, S25 and GaAs can vary significantly. The
graphs presented in Fig. 9 refer to tubes considered as typi
cal based on practical experience of author.
Differences between different generations of IITs are
much deeper that the simple scheme shown in Fig. 9. Sev
eral more parameters should be considered in order to show
important differences between tubes of different genera
tions. Photocathode material is only one of these parame
ters. Next, the differences between different generations are
not always clear. Further on, some of generations can be fur
ther divided into subgenerations. All these subtle details
shall be now discussed.
Generation 0 refers to the technology developed during
World War II, employing fragile, vacuumenveloped image
converters with poor sensitivity and little gain. These were
single stage tubes that achieved image intensification due to
acceleration by high voltage of electrons emitted by the
photocathode and striking the phosphor screen. S1 (sil
veroxygencaesium) photocathode, electrostatic inversion
and electron acceleration were typically used to increase
brightness of input image.
S1 photocathode has two small peaks of its sensitivity:
the first in ultraviolet (UV) and the second in near infrared
(NIR) about 800 nm, but is characterized by low sensitivity
at visible band (Fig. 9). This situation fits badly for a task to
amplify brightness of input image created by nocturnal light
characterized by negligible amount of UV light and high
amount of visible light. Therefore, luminous sensitivity of
S1 photocathodes was not higher than about 60 A/lm
(microampere per lumen). Further on, luminance gain was
no more than about 150 lm/lm. Such a low luminance gain
is not sufficient to create a bright image of scene of interest
at typical night conditions. Therefore, Gen 0 tubes were in
the past used in active night vision systems cooperating with
an IR illuminator. High power tungsten bulbs covered with
an IR filter suppressing visible radiation were used as illu
minators. Active character of use of these first night vision
devices was their significant disadvantage.
Generation 1. First Gen 1 tubes were in general im
proved Gen 0 tubes. Initial experiments with new photo
cathode materials showed that S11 photocathode (cesium
antimony) is characterized by an extremely high quantum
efficiency up to 20%) but only in a visible range. Therefore,
only slight improvement of luminous sensitivity to the level
of 80 A/lm was achieved using this new type of photo
cathodes because the value of luminous sensitivity depends
more on tube sensitivity in near infrared band than on
sensitivity in visible band.
The breakthrough came about 1956 with a discovery of
S20 photocathode (multialkali photocathode: sodiumpo
tassiumantimonycesium) that is sensitive in both visible
and near infrared (Fig. 9). Significantly improved photoca
thode sensitivity (luminous sensitivity up to 200 A/lm),
and improved technique of electrostatic inversion and elec
K. Chrzanowski
157

tron acceleration enabled to achieve luminance gain from
about 400 lm/lm to about 800 lm/lm. Because of this quite
high luminance gain some of Gen 1 NVDs were used as
passive night vision systems, but majority of Gen 1 tubes
was still used in active systems. The reason for using sup
port of artificial infrared illuminators is the fact that much
higher luminance gain in the order over 30 000 times is nec
essary to achieve ability to see even at medium illuminated
(overcast quoter moon) night conditions.
First generation IITs are characterized by good image
resolution (2530 lp/mm), a wide dynamic range (the ability
to reproduce the ratio between bright and dark parts of an
image), low noise, and clear image with few blemishes. Due
to earlier mentioned advantages and low production costs,
Gen 1 tubes are still manufactured and NVDs built using
Gen 1 tubes still dominate in commercial market. However,
Gen 1 tubes are only very rarely used for military applica
tions due to low luminance gain and significant distortion
present in images generated by these tubes.
From technical point of view Gen 1 tubes are apparently
simple devices. Focusing is achieved by using an electron
lens to focus electrons, originating from the photocathode,
onto the screen (Fig. 10). In the inverter diode tube pre
sented in Fig. 10 an electrostatic field directs the photoelec
trons and focuses an inverted image on the phosphor screen.
Electron lens can be achieved by combining an electrosta
tical field with an axial magnetic field provided by either
a solenoid or permanent magnet. A uniform magnetic field
enables to achieve a good resolution over the entire screen
and at the same time keeps distortion low. Fibre optics win
Fig. 10. Gen 1 tube: (a) diagram of a NVD built using a Gen1 tube,
and (b) photo of Gen 1 tube with external electronic power supply.
158
dows are used in Gen 1+ tubes to minimize degradation of

the image resolution towards the edge of the tube. The fibre
optics can potentially enable also efficient coupling to
another image tube, to an imaging detector, or to photogra
phic film.
As it was earlier said, luminance gain of a single Gen 1
tube was still too low to enable design a truly passive
NVDs. A simple solution to overcome this drawback was
proposed in a form of a cascade tube built by combining two
or three single tubes. The experiment carried out in 1950s
showed that it is possible to design a cascade tube of lumi
nance gain even over the level of 30 000 lm/lm capable to
produce a usable image of scenery of interest even at star
light conditions. The first cascade tubes coupled using opti
cal systems were too bulky (length up to about 40 cm) to be
used in NVDs for military applications. However, introduc
tion of fibre optics for coupling of image intensifier tubes
enabled designing of much shorter cascade tubes that were
used in big numbers in devices manufactured during Viet
nam war in 1960s and partially also during 1970s.
Two other big drawbacks of the cascade tubes were low
SNR and high cost of manufacturing. The first drawback
means that output images were very noisy because the next
stage of the tube amplified noise present in the image gener
ated by the previous stage. The second drawback was cau
sed by the necessity to use typically three single stage tubes
to design a single cascade tube. Because of these three ear
lier discussed drawbacks the cascade tubes were quickly
eliminated with advent of Gen 2 tubes in 1970s. Nowadays
the cascade tubes are only rarely used in scientific appli
cations.
Generation 2 image intensifier tubes represent a signifi
cant breakthrough in night vision technology. These are
small, compact IITs that offered luminance gain at the level
of about 30 000 lm/lm and the later even more. Such a sig
nificant increase of tube luminance gain, while making tube
also smaller, was achieved due to four basic reasons.
First, Gen 2 tubes use microchannel plate (MCP) to
amplify electrons emitted from photocathode (Fig. 11). The
MCP is a very thin plate of conductive glass containing mil
lions of small holes. An electron entering a channel strikes
the wall and creates additional electrons, which in turn cre
ate more electrons (secondary electrons), again and again
(Fig. 11). The microchannel plate is an array of miniature
electron multipliers oriented parallel to one another and
have length to diameter ratios between 40 and 100. Channel
axes are typically normal to, or biased at a small angle ( 8)
to the MCP input surface. The channel matrix is usually fab
Fig. 11. Principle of work of a microchannel plate.
2013 SEP, Warsaw
ricated from a lead glass, treated in such a way as to opti

mize the secondary emission characteristic of each micro
channel and to render the channel walls semiconducting, so
as to allow charge replenishment from an external voltage
source. Parallel electrical contact to each channel is pro
vided by the deposition of a metallic coating on the front
and rear surfaces of the MCP, which then serve as input and
output electrodes, respectively. The total resistance between
electrodes is in the order of 109 W. Such microchannel
plates allow electron multiplication factors of 103105 de
pending on channel length and number of layers. Spatial
resolution is limited by channel dimensions and spacing; 12
m diameter channels with 15 mm centertocenter spacings
was typical for first and second generation tubes. Nowadays
both dimensions and spacing can be about two times
smaller.
Second, Gen 2 tubes use new S25 photocathode. This
is actually the same multialkali photocathode as S20 pho
tocathode used in Gen 1 tubes but S25 photocathode is
built using thicker layers of the same materials. In this way
extended red response and reduced blue response was ach
ieved making sensitivity spectrum of S25 photocathode
well matched to spectrum of nocturnal light. Luminous sen
sitivity of the first S25 photocathodes was about 250

A/lm. It is a noticeable improvement in comparison to
S20 photocathode used in Gen 1 tubes but almost negligi
ble in comparison to revolution in luminance gain achieved
by introduction of MCP plate.
Third, first Gen 2 tubes were inverter tubes that used an
electrostatic inversion, in the same way as the Gen 1 tubes
did, but with an added MCP (Fig. 12). The size of Gen 2 in
verter tubes was only slightly lower than size of Gen 1 tubes.
However, introduction of proximity focused tubes that could
also carry out an image inversion using a fibre bundle with
a 180 degree twist in it enabled the design of much smaller
tubes (Fig. 13). Nowadays, almost all tubes used for night vi
sion applications are proximity focused tubes.
Fourth, Gen 2 tubes were integrated with small elec
tronic modules capable not only to power tubes using high
voltage but also to carry out automatic gain control and
bright spot protection. This new feature is a sharp contrast to
a bulky external high voltage power supplies used by Gen 1
tubes (Fig. 10).
To summarize, Gen 2 tubes should be treated as a real
revolution in image intensification technology. There is
a big difference between Gen 2 tubes and Gen 1 tubes in sit
uation when border between Gen 3 tubes and Gen 2 tubes is
much more blurred, especially between Gen2+ tubes and
Gen 3 tubes.
Fig. 12. Inverter Gen 2 tube: (a) diagram and (b) photo (tube dimen
sions F52.855.5 mm).
Fig. 13. Proximity focused Gen 2 tube: (a) tube diagram and (b)
photo (tube size F35.531 mm).
K. Chrzanowski
159

Two plus generation (Gen 2+) is not formally recog
nized by U.S. authorities. This term refers to technologies
used to improve sensitivity of the trialkali S25 photoca
thodes and manufacturing new generation microchannel
plates. In detail, photocathode sensitivity was improved
more than two times, microchannel plates of highly increa
sed openarea ration up to about 70%, and reduced internal
noise were developed. These new technologies enabled to
manufacture Gen 2+ tubes of expanded spectral sensitivity
up to about 950 nm and of luminous sensitivity up to about
700 A/lm. First Gen 2+ tubes were developed by Photonis
company (France) in 1989. Further development of Gen 2+
tubes has been continued by several nonU.S. manufactur
ers in France, Netherlands, Russia, China, and India. Nowa
days Gen 2+ tubes (marketed using different names like
Gen II Plus, SuperGen, HyperGen, XD4, XR5) represent
majority of hiend image intensifier tubes manufactured by
nonU.S. manufacturers.
Gen 3 tubes are very similar to the Gen 2 tubes from
design point of view. The primary difference is the material
used for the photocathodes. The second generation image
intensifiers use photocathodes with a multialkali coating
whereas the third generation image intensifiers uses photo
catodes with a GaAs/GaAsP coating. The latter photoca
thodes are characterized by higher sensitivity and addition
ally the spectral sensitivity band is extended more in near
infrared (Fig. 9).
Manufacturing of Gen 3 tubes was started in 1980s but
some development works were done earlier in 1970s. Liter
ature sources state that the photocathodes used in Gen 3
tubes are characterized by radiant sensitivity in the near
infrared about three times better in comparison to photo
cathodes with a multialkali coating used by Gen 2 tubes
(Fig. 14) [8].
Higher radiant sensitivity of Gen 3 tubes means higher
luminous sensitivity. SNR is theoretically proportional to
square root from luminous sensitivity. Therefore, the situa
tion shown in Fig. 14 should theoretically guarantee SNR of
Gen 3 tubes to be about two times better than for Gen 2

tubes. However, such a conclusion is typically not true due
to two main reasons. First, radiant sensitivity of modern
Gen 2+ tubes is much better than data shown in Fig. 14. Sec
ond, Gen 3 photocathodes have a significant drawback.
These photocathodes can be quickly seriously degraded by
positive ion poising that can reduce photocathode sensitiv
ity up to about two times within a period about 100 hours. In
order to protect the photocathode from positive ions and
gases produced by the MCP, manufacturers of Gen 3 tubes
have added a thin film of sintered aluminium oxide attached
to entrance of the microchannel plate. This technique pro
tects effectively photocathode but the protecting film traps
about half of electrons emitted by the photocathode. These
trapped electrons will not be amplified. Therefore, the effec
tive luminous sensitivity of Gen 3 tube can be almost two
times lower than photocathode luminous sensitivity of such
a tube. The final result is that SNR of typical Gen 3 tubes is
often comparable to good Gen 2+ tubes. The diagram pre
sented in Fig. 15 explains graphically the earlier mentioned
effect of reduction of effective sensitivity of Gen 3 tubes.
The protecting film generates also significant blurring
that occurs when it is bombarded by a high intensity con
centrated electron beam. Therefore, halo effect in Gen 3
tubes is typically significantly bigger than in Gen 2+ tubes
(Fig. 16).
In 1998 the U.S. company Litton informed general pub
lic about the development of the filmless image intensifier
tube built using the GaAs/GaAsP photocathodes [10]. De
velopment of such tubes meant elimination of main draw
back of Gen 3 tubes the ion protecting film trapping elec
trons emitted by photocathode (Fig. 17). This new techno
logical solution could potentially increase SNR of IITs at
least over 25%. Therefore, the new tubes caused significant
Fig. 14. Radiant sensitivity curves of Gen 2 tube photocathode

(S25 extended red multialcali) and Gen 3 tube (GaAs) photoca
thode (after Ref. 8).
Fig. 15. Two tubes (Gen 2+ tube XD4 class and Gen 3 tube) of
much different luminous sensitivity generating output image of the
same signal to noise ratio (after Ref. 13).
160
2013 SEP, Warsaw
Fig. 16. Image generated by two night vision devices built using im
age intensifier tubes of different halo: (a) big halo, and (b) small halo
(after Ref. 9).
interest of U.S. military authorities and Night Vision and

Electronic Sensors Directorate (NVESD) assigned these
new tubes as Gen 4 tubes. However, it was soon found that
the new Gen 4 tubes are too fragile for real military condi
tions (sensitivity to mechanical shock) and additionally can
not pass typical reliability tests when the tubes are exposed
to sudden uniform/concentrated flashes of light. Therefore,
in 2002 NVESD revoked its previous decision designating
filmless tubes as Gen 4 tubes. Since that time these new
tubes are called as Gen 3 Filmless tubes and there is offi
cially no Gen 4 tubes.
Despite of setback with reliability problems develop
ment of Gen 3 filmless tubes have been continued and
brought significant technology improvements. Two tech
niques have been found to overcome the problem of photo
cathode poisoning. First, using of improved scrubbing tech
niques during manufacture of the microchannel plate that is
the primary source of positive ions in a wafer tube. Second,
using of autogating technique to power photocathode beca

use a sufficient period of autogating causes positive ions to
be ejected from the photocathode before they could cause
poisoning of the photocathode [10].
Gen 3 Filmless tubes are characterized by excellent SNR
(can be even over 30) and can produce clear image of scen
ery of interest even under very dark, moonless nights. The
refore, Gen 3 Filmless tubes are often used in night vision
goggles for aviators or for special operation teams, but are
avoided in night vision sights due to their vulnerability to
mechanical shock.
Gen 3 Filmless tubes represent one group of signifi
cantly improved Gen 3 tubes. Gen 3 Thin Film tubes repre
sent another group of improved Gen 3 tubes, and are often
called Gen 3+ tubes.
Gen 3 Thin Film technology was developed by ITT Night
Vision (the biggest U.S. manufacturer) as a response to com
petition from Gen 3 Filmless tubes offered by another U.S.
manufacturer Litton. ITT Night Vision found that by signif
icantly thinning, rather than removing the protective film, it
could achieve the armymandated Gen 4 performance and
endoflife reliability requirements [12]. Maintaining the film
also would protect all the important gallium arsenide pho
tocathode structure. Therefore, Gen 3 Thin Film tubes use ion
protecting ultrathin film of the thickness about 3 nm (typical
situation) or sometimes as thin as 1 nm. Reduced voltage ap
plied to photocathode is another but minor change in
comparison to typical Gen 3 tubes.
Thin Film technology is not as effective as filmless tech
nology in eliminating trapping of electrons emitted by pho
tocathode. Up to 25% of electrons are still trapped by the
Fig. 17. Graphical presentation of difference between Gen 3 tubes and Gen 4 tubes: (a) typical Gen 3 tubes, and (b) filmless Gen 4 tube (after
Ref. 11).
K. Chrzanowski
161

thin protecting film. However, these losses are much lower
than in case of typical Gen 3 tubes and Gen 3 Thin Film
tubes offer very good SNR; almost as high as for Gen 3
Filmless tubes. At the same time Gen 3 Thin Film tubes are
characterized by better reliability than Gen 3 Filmless tubes.
Therefore, Gen 3 Thin Film tubes are nowadays the main
group of hiend image intensifier tubes used by U.S.
military.
Short summary of basic features of different generations
of IITs is shown in Table 1. It is clear that the generation
number is a code that gives general description about tech
nology that is used to manufacture the tubes. There are
subclasses within each of main generations depending on
technology details.
The presented earlier division of image intensifier tubes
into several generations is based on a manufacturing tech
nology as the main criterion. As it was mentioned earlier
there is no clear precision relationship between generation
number and tube performance.
Gen 3 Filmless tubes and Gen 3 Thin Film tubes are two
technologies that significantly improved performance of
image intensifier tubes within last decade. Gating is the
third less noticeable technology that improved performance
of modern tubes.
Gating means a technique used to switch on and off
image intensifier tube like an electronic gate. For several
decades this technique has been used in gated image intensi
fiers used in high speed imaging systems to enable visual
ization of ultra fast temporal events (time intervals in nano
seconds or even picoseconds). More recently gated tubes
have found application in active night vision systems that
can potentially enable visualization of targets behind semi
transparent obstacles (including fog) and to measure dis
tance to the target. However, both high speed gated night
vision systems and active gated night vision systems are
rather exotic types of night vision systems that are used
rather rarely.
Autogating is a type of gating technique that has found
mass application in modern image intensifier tubes. The
experiments with this technique were carried out for a de

cade or more, but autogated tubes appeared on market about
2006 year. These tubes gate voltage applied to photocathode
and MCP in order to keep constant current flowing through
the MCP even when tube is strongly illuminated. The gating
is done at very high frequency and is not noticeable on out
put image. Autogated tubes can be operated at much bri
ghter conditions than typical tubes without damaging tubes
or blurring output image. In other words, the autogated
tubes are characterized by ultra extended dynamic that the
tubes can generate clear image in both dark nights and twi
light (or even day time) conditions.
In detail, resolution of typical tubes can drop several
times when tube operates at day level illumination (from
about 60 lp/mm to about 1020 lp/mm). Resolution of auto
gated tubes drops at the same conditions no more than at
about 20% (from about 60 lp/mm to about 50 lp/mm).
Therefore, this new feature is of high value in military/secu
rity applications where illumination conditions can change
very rapidly within several seconds and users of NVDs built
using nonautogated tubes are blinded by sudden flashes of
light or get blurred image at day conditions. At present,
autogated tubes becomes increasingly popular in hiend
tubes for military applications. However, it should be noted
that autogating technique can be used in both Gen 2+ tubes
and Gen 3 tubes and the term autogating has no direct
connection with a particular generation of image intensifier
tubes.
3.2. Tube performance

There is a common view that higher generation number
means a better tube. It is true if we compare Gen 0, Gen 1 or
Gen 2 tubes but do not have to be true if we compare Gen 2
and Gen 3.
Main EU manufacturer of image intensifier tubes, Pho
tonis, claims that generation numbering based on manufac
turing technology introduced by U.S. laboratories cooperat
ing with U.S. industry is misleading [13]. Therefore, Pho
Table 1. Basic parameters of IITs from different generations.

Gen
No.
Photocathode
material
Photocathode
sensitivity
[A/lm]
Design type
Luminance gain
[lm/lm]
Resolution
[lp/mm]
SNR
S1
< 60
inverter tube
< 200
2060
S20
< 160
inverter tube
< 800
2060
1+
S20
< 160
cascade inverter tube
< 20 000
2030
58
S25
< 350
inverter MCP tube
< 50 000
2443
1217
2+
improved S25
< 700
proximity focus MCP tube
< 70 000
4381
1524
GaAs/GaAsP
< 1600
proximity focus MCP, tube

with protecting film
< 70 00
3664
1825
3+ Thin Film
GaAs/GaAsP
< 1800
proximity focus MCP, thin

film tubes
< 70 000
5771
2428
3 Film less
GaAs/GaAsP
< 2200
proximity focus MCP,

filmless tubes
< 80 000
5771
2431
162
2013 SEP, Warsaw
tonis (formerly also DEP) introduced its own generation

numbering. The tubes are divided into four generations:
1. Gen II Plus;
2. SuperGen;
3. XD4;
4. XR5.
This division is based on tube performance character
ized by two important parameters: resolution and SNR, or
on the product of these two parameters called figure of merit
(FOM). Typical values of resolution, SNR and FOM param
eters used by Photonis to characterize mentioned above gen
erations of image intensifier tubes are shown in Table 2
[13].
Table 2. Division of image intensifier tubes proposed by Photonis.
Generation code
Gen II Plus
SuperGen
XD4
XR5
Resolution [lp/mm]
36
51
64
72
SNR
13
21
23
28
FOM
468
1071
1472
2016
*Values of resolution and SNR are typical values from data sheets
at Ref. 13.
The division of image intensifier tubes into several per

formance generations presented in Table 2 should be treated
as a division used locally by only one of main manufactur
ers. However, the general concept to classify tubes into sep
arate groups on the basis of values of two important parame
ters like resolution and SNR is quite popular.
U.S. Department of State guidelines for export of image
intensifier tubes are nowadays based on FOM values. It is
highly probable that the European Community (EU) author
ities will follow this example and the old guidelines based
mostly on photocathode luminous sensitivity will be re
voked. A big advantage of FOM criterion is the fact that the
FOM parameter of typical potted tubes can be measured
without any damage to the tube in situation when measure

ment of luminous sensitivity is destructible to potted tubes.
Next, U.S. authorities generated division of image inten
sifier tubes into generation based on manufacturing technol
ogy as the main criterion. According to typical logic tubes
of higher generation should be better than tubes of lower
generation. Therefore, this official division discussed in
Section has been used by U.S. manufacturers to promote
U.S. made Gen 3 tubes against Gen 2+ tubes made almost
exclusively by nonU.S. manufacturers. However, as it was
shown earlier there is no clear precision relationship bet
ween generation number and tube performance. Therefore,
even in USA, requirements for big purchase programs (co
ded Omnibus) carried out cyclically by US military have
been based mostly not on manufacturing parameters but on
performance parameters like resolution and SNR. There
fore, tubes manufactured by U.S. manufacturers are often
divided into generations using requirements of Omnibus
programs (up to Omnibus VII). Technical requirements on
MX 10160 image intensifier tube purchased within Omni
bus programs are presented in Table 3 [14]. As we see in
this table performance of tubes of the same MX 10160
family purchased within different Omnibus program differ
very significantly.
It was reported recently that contracts within new Omni
bus program (OMNI VIII) were awarded [15]. However,
technical details of OMNI VIII are not known and will not
be discussed here.
To summarize, tube performance is not directly related
to generation number. Situation when Gen 2+ tubes is better
than Gen 3, or even to a lesser degree better than Gen 3+ is
possible but inverse situation is equally probable. All tube
manufacturers claim advantages of their tubes but often for
get about drawbacks.
There are three technologies that are competing in mar
ket of hiend tubes for ultra demanding military applica
tions: improved Gen 2+ tubes (XR5 type or equivalents),
Table 3. Requirements on MX 10160 type image intensifier tube in Omnibus programs.

Contract
Resolution lp/mm
Omni I
Omni II
Omni III
Omni IV
Omni V
Omni VI
Omni VII
36
45
51
64
64
64
64
S/N
16.2
16.2
19
21
21
25
28
FOM
583
729
969
1344
1344
1600
1792
Photocathode sensitivity
A/lm@2856 K
1000
1000
1350
1800
1800
2000
2200
Gain [fL/fc]
[email protected] lp/mm
2000035000 4000070000 4000070000 4000070000 4000070000 5000080000 5000080000

0.83
0.83
0.9
0.92
0.92
0.92
0.92
[email protected] lp/mm
0.58
0.6
0.7
0.8
0.8
0.8
0.8
MTF@15 lp/mm
0.28
0.38
0.45
0.61
0.61
0.61
0.61
MTF@25 lp/mm
0.08
0.18
0.2
0.38
0.38
0.38
0.38
Halo (mm)
1.47
1.47
1.47
1.25
1.25
0.90
0.70
Phosphor
P20
P20
P20/P43
P43
P43
P43
P43
Year
1982
1985
1990
1996
1999
2002
2006
K. Chrzanowski
163

Gen 3 Thin Film tubes, Gen 3 Filmless tubes. As we see in
Table 4 each of these technologies has its strong points and
weak points.
Table 4. Comparison of three competing technologies of hiend
IITs.
Gen 3
Gen 3
Thin Film
Filmless
Resolution
81 lp/mm
64 lp/mm
64 lp/mm
SNR
26
28
30
Reliability
15000 hours
15000 hours
10000 hours
Very high High resolution, high Very good
Advantages
resolution SNR, high reliability
SNR
Medium
No
clear
superior
Modest
Disadvantages
SNR
parameter
reliability
Gen 2+
*Attention: Values presented in the table refer to parameters of

best tubes manufactured using compared technologies. Typical
values are often lower.
As it was mentioned earlier the concept to characterize

IITs using FOM criterion is becoming increasingly popular
because FOM is directly related to tube performance. How
ever, the readers should be warned that there is a significant
drawback of this criterion: low accuracy of measurement of
both resolution and SNR used to calculate FOM.
Low accuracy of measurement of resolution of IITs is
caused by several factors: big steps (13%) between different
resolution patterns of typical USAF 1951 resolution target
used for resolution measurement, human subjective error,
and influence of projection and observation system. It was
reported by one of tube manufacturers that error of the mea
surement of resolution is often more than 10lp/mm [16]. In
author opinion there are ways to minimize error of resolu
tion measurement below mentioned above level, but still the
error is noticeable.
Measurement error of SNR is often even bigger due to
high technological challenges on light meters used for mea
surement procedure and slight differences between mea
surement procedures used by different manufacturers. The
accuracy limits illuminance/luminance meters needed for
SNR measurement are acknowledged by MIL standards that
require only modest 10% accuracy of such meters [17].
Due to these reasons differences between FOM values of
the same tube measured by two test teams at the level of
20% are often met. Therefore, FOM values presented by
different manufactures and measured by different teams
using different apparatus should be treated cautiously. How
ever, in spite of this limitation FOM criterion should still be
treated as much better solution to characterize tube per
formance than generation number.
tapers. The fibre taper was in direct contact with the photo
cathode and transmitted here an image created by the objec
tive at input plane of the fibre taper (Fig. 18a). Nowadays
tubes with input fibre optics are rarely met. Modern tubes
are built using input glass optics (typical quartz, borosilicate
glass, synthetic silica or rarely MgF2). In the latter case the
objective must create image at photocathode plane as shown
in Fig. 18(b).
The difference between NVDs built using tubes having
different input optics goes however much deeper than only
difference in plane where the image is to be focused (typi
cally about 5 mm difference). Two more important factors
should be taken into account.
1. Fibre optics is characterized by almost two times worse
transmission in comparison to transmission of input
glass optics due to presence of dead areas in fibre tapers.
The consequence is much lower luminous sensitivity of
image intensifier tubes built using input fibre optics be
cause photocathode of such tubes gets less photons than
photocathode of a tube built using input glass optics at
the same illumination conditions.
2. Size of fibre optics creates limit on maximal resolution
of an image that can be transmitted.
3.3. Tube input optics

Photocathode of image intensifier tube must be protected
from direct contact with atmosphere by some kind of input
optics. First proximity focus image intensifier tubes (most
common type of Gen 2 and Gen 3 tubes) used fibre optics
164
Fig. 18. Image creation process in NVDs: (a) case of a tube with in
put fibre optics, and (b) case of a tube with glass optics.
2013 SEP, Warsaw
Because of these two discussed reasons the tubes built

using input fibre optics are nowadays very rarely met and
the tubes built using input glass optics are typically used in
modern NVDs.
So far, only advantages of input glass optics were men
tioned. It should be, however, also noticed that the tubes
built using input glass optics create significant problems for
designers of optical objectives for NVDs. The reason is
a simple difference in location of image created by the
objective (Fig. 18).
In case of NVDs built using tubes having fibre optics,
the task of the objective is to create an image at input plane
of the fibre optics plate. The fibre optics is as an external
module not related to the optical objective.
In case of NVDs built using tubes having glass optics the
task of the objective is to create an image at exit plane of the
glass optics plate. Now, the glass optics becomes an internal
part of the objective. Next, the glass optics distorts transmit
ted image and this distortion must be corrected by design of
the objective.
The distortion added by the glass optics is particularly
strong in case of ultrabright objectives having low Fnum
ber (ratio of focal length to aperture). Therefore, design of
aberrationfree night vision objectives having Fnumber
below one for NVD built using a tube with input glass
optics is considered as a real challenge for optical designers.
Upgrading of old NVDs built using IITs having input
fibre optics to modern devices built using tubes having input
glass optics by simple tube replacement is usually not possi
ble because of different requirements on optical objective.
Replacement or modernization of an old optical objective is
typically needed, as well.
3.4. Tube output optics

Tube output optics that covers phosphor screen of image
intensifier tubes is needed for two reasons. First, to protect
the screen from direct contact with atmosphere. Second, to
invert image created by the screen. The latter function is
needed in great majority of NVDs built using noninverting
oculars to prevent situation that the user will see inverted
image of observed targets (Fig. 19).
There are two main types of output optics: glass optics
and fibre optics. The tubes with output glass optics do not
invert images but such tubes are cheaper than tubes with
output fibre optics.
Low cost tubes with output glass optics (plate made
from quartz, or borosilicate glass) are used in applications
where image orientation is not important (like astronomy)
or image created at tube screen is inverted using special
inverting oculars or other optical systems.
More expensive tubes with output fibre optics are used
in great majority of NVDs. Such tubes are offered in two
versions: tubes with inverting fibre optics and tubes with
straight fibre optics. The inverting fibre optics is used in
tubes for most of NVDs. Tubes with straight fibre optics are
Fig. 19. Image inversion process: (a) NVD built using a tube with
output glass optics, and (b) NVD built using a tube with output in
verting fibre optics.
rarely used in NVDs built using non typical optical systems

(noninverting objective or inverting ocular).
The output surface of the fibre optics is typically curved
[(Fig. 19(b)]. The are two basic reasons to use such curved
fibre tapers. First, it is easier to design aberrationfree ocu
lars for curved input plane than for flat input plane. Second,
image of phosphor screen generated via curved fibre taper
and seen by an ocular is more uniform for curved fibre
tapers than in case of flat fibre tapers. In the latter case
image of the centre looks significantly brighter than image
of the outer sectors.
3.5. Phosphor screen

Phosphor screen is one of most important parts of an IIT.
This part is used to convert the electrons, emitted by the
photocathode and multiplied by the MCP incoming to the
phosphor screen (the anode), into light, visible for human
observer. The screen is typically a fibre optic plate coated
with a phosphor and additionally coated with an external
layer of aluminium. This layer of aluminium is needed as
electrical contact to the phosphor layer using metal contact
rings. The aluminium layer increases light efficiency up to
almost 100% because light created by the phosphor is re
flected back by the aluminium layer. The conductive alu
minium layer reduces also the electrical charge from
electrons and ions.
An electron acceleration energy of about 3 kV is re
quired to pierce the aluminium layer. If this acceleration
K. Chrzanowski
165

voltage is too high for a particular application then indium
tinoxide (ITO) coated screens are used to provide the con
ductive base needed to minimize the electrostatic effects of
charged particles [18].
Phosphors are made of rare earth oxides or halides (e.g.,
gadolinium, lanthanum, yttrium). Most phosphors are clas
sified by using P letter and suitable number (from P1 to
about P56). There are many known phosphors manufac
tured for cathode ray tubes (CRT tubes) that could be poten
tially used also for screens of IITs. However, two phosphors
are used in majority of modern tubes: P20 and P43. There
are also several other phosphors that are also used but much
more rarely: P45, P46, and P47.
Table 5. Basic parameters of typical phosphors used in screens of
IITs.
No
1
2
3
4
5
Luminous
Phosphor Phosphor decay
Colour
efficiency
type
time (down to 1%)
P20
60 ms
yellowgreen
high
P43
3 ms
yellowgreen
high
P45
5 ms
white
moderate
P46
3 s
green
low
P47
0.4 s
blue
very low
*P22 phosphor of similar parameters like P20 phosphor is also

used.
**Parameters of phosphors depends on many factors. Data shown
in Table 5 should be treated only as exemplary parameters.
There are four critical parameters of phosphors for ima

ge intensifier tubes: conversion factor, light spectrum (emit
ting colour), time decay, and resolution.
Typical conversion factors lie between 20 and 200 pho
tons per electron, depending on the phosphor and the kinetic
energy of the electrons, e.g., the acceleration voltage [19].
High conversion factor is not enough to guarantee that
the screen shall produce a bright image perceived by human
observer. The same number of photons can create totally
different perception of brightness depending on wavelength
of emitted light. Due to this reasons phosphors that emit
light at greenyellow colour, where eye sensitivity is the
highest, are preferred. Next, phosphors for image intensifier
technology are typically evaluated using luminous efficien
cy as the evaluating criterion. The latter parameter is defi
ned as a ratio of output luminance to number o input
electrons.
Phosphor decay time is typically defined as time interval
needed for phosphor screen to achieve low relative screen
brightness (about 1%) from the moment when the input
light is switched off. In case of tubes for night vision appli
cations it is typically required that the phosphor decay time
should be at least several times shorter than temporal inertia
of human eye usually considered to be in 100200 ms
range. As we see in Table 5 there can be some doubts if P20
phosphor fulfils this criterion.
Resolution of phosphors, or in other words phosphor
ability to create very fine images, depends mostly on me
dian particle size. The latter parameter can vary from about
166
2 m to about 12 m. Bigger particles makes possible to

built screens of higher conversion factor but phosphor reso
lution is reduced. In general manufacturers of image intensi
fier tubes try to keep resolution of the phosphor at the level
about two times better than resolution of final image intensi
fier tubes. In practice it means that resolution of phosphor
screen should be in the region from about 80 lp/mm to about
160 lp/mm.
Nowadays, image intensifier tubes built using P20/P43
phosphors dominate on market due to high luminous effi
ciency. Wavelength of maximal spectral emission of P43
(545 nm) nearly perfectly fits to maximal spectral sensitivity
of human eye (555 nm) and high luminous efficiency this
phosphor is the effect of this near perfect fit (Fig. 20). High
spectral efficiency of P20 phosphor is a result of wide spec
trum of P20 phosphor that almost fully overlap spectral sen
sitivity of human eye sensitive mostly in band from about
500 nm to 600 nm.
The P20 phosphor has been the preferred solution for
several decades. However, it is also characterized by rela
tively long decay time that makes this phosphor less suitable
for applications where short temporal inertia is important.
Therefore, nowadays P43 phosphor is typically used in
tubes targeted for airborne applications or other applications
where night vision devices are supposed to be used for sur
veillance of dynamic phenomena.
P45 phosphor is a novelty in night vision technology
introduced on the market by one of big manufacturers of
IITs Photonis within the last several years. It is adver
tised that P45 phosphor makes images generated by image
intensifier tubes more natural for humans who see targets at
night in gray scale not in colours [13]. It is true that humans
see images under dark conditions as monochromatic images
of different gray level. Next, users of typical night vision
goggles emitting greenish images see images in false
colours for some time after goggles are removed from their
eyes. The latter effect is eliminated when using goggles
emitting blackwhite images.
Fig. 20. Emission spectra of several phosphor screens (after

Ref. 20).
2013 SEP, Warsaw
However, in spite of advantages mentioned above, there

are serious doubts whether P45 makes image intensifier
tubes any better due to several reasons.
Images generated by P45 screens look darker than im
ages generated by typical P20/P43 screens for the same
tube design due to lower luminous efficiency of P45
phosphor in comparison to typical P20/P43 phosphors.
Therefore, luminance gain of tubes built using P45
screens is often almost two times lower than luminance
gain of similar tubes built using P20/P43 screens.
Tubes with P45 screens emit radiation at entire visible
band including red/orange spectrum. Red/orange light
sources are prohibited in airborne applications.
Humans looking into greenish screens of IITs quickly
adapt to monocolour vision and start seeing in gray
scale like during naked eye vision at night conditions.
Human observes are less tired when looking on tubes that
generate greenyellow images in comparison to tubes gen
erating whiteblack images. The latter effect is because
light spectrum of P20/P43 tubes better fits to eye spectral
sensitivity curve than light spectrum of P45 tubes.
It is easier to design optical oculars optimized for narrow
spectral band of P43/P20 phosphors than for much wider
spectral band of P45 phosphor.
Because of the reasons mentioned above the blackwhite
images should not be treated as an advantage over classical
greenyellow images. Inverse conclusion is more reason
able. Probability that P20/P43 screens in IITs shall be re
placed in near future by P45 tubes is rather low. The same
can be said about P46/P47 phosphors that are mostly used in
some scientific applications where short temporal inertia of
phosphor screen is crucial but are rarely used in main stream
technology of night vision.
l
The weight of the 16 mm tube has been reduced by 35

grams compared to standard ANVIS 18 mm tube while its
volume size has been reduced by 40%. The 16 mm tube is
available with an integrated autogated power supply [13].
The size difference in comparison to typical 18 mm tube
mm tubes is not big. Therefore, the performance differences
between new 16 mm tubes and typical 18 mm tubes are not
big, too. It is risky to predict if the new size is accepted by
other night vision manufacturers and if 16 mm tube be
comes a new standard for applications where size and
weight are of crucial importance. However, some manufac
turers of NVDs have already started production of ultra
weight night vision monoculars using these new ultra small
tubes [22]. Therefore, it can be rather expected that most
manufacturers of IITs will soon offer these new miniaturi
zed 16 mm tubes.
Fig. 21. Images of the same scene generated by two tubes with dif
ferent phosphor: (a) typical XR5 tubes with P43 phosphor, and
(b) Onyx tube with P45 phosphor (after Ref. 21).
3.6. Photocathode diameter

All types of IITs are also differentiated by the nominal use
ful diameter of the photocathode. Typical diameter values
are 18 mm, more rarely 25 mm. Tubes of bigger sizes are
used sometimes in scientific applications.
However, recently a new 16 mm image intensifier tube
was introduced by one of main manufacturers of IITs in
order meet new requirements for reduced size, weight, and
minimal power consumption [13].
4. Optics of NVDs
Image intensifier tube is rightly considered as the most
important module of night vision devices. Because of this
unquestionable importance of the tubes there is a tendency
to decrease design complexity of optical modules of NVDs
and influence of optical modules on overall performance of
NVDs.
4.1. Design requirements

The night vision optical modules (objective, ocular) looks
apparently simple and similar to low cost optics used in bin
oculars, close circuit television (CCTV) cameras, and pho
tography. Two optional modules (beam splitter and optical
filter) look also apparently simple. However, practically
design of high performance night vision optical modules is
a technological challenge due to a set of different reasons.
4.1. 1. Objective
There are five main requirements on objectives for modern
NVDs:
1. Field of view (FOV) optimized for type and applications
of NVD. Typical FOV of night vision goggles and
monoculars is about 40 but objectives of bigger FOV
up to about 50 are sometimes used, too. Objectives of
FOV in the range from about 4 to about 12 are used in
night vision sights or binoculars. Possibility to change
FOV of the objective by adding additional lens is also
sometimes required.
2. Low Fnumber. It is typically required to design objec
tives of Fnumber (ratio of focal length to aperture)
lower than 1.2. However, objectives of Fnumber equal
to 0.9 or lower are used, as well. It should be noted that
that Fnumber of photographic objectives is rarely lower
than 1.8.
3. Optimization for wide spectral band. Objectives for
NVDs should be optimized for spectral band from about
500 nm to about 900 nm. This is a significant increase in
requirements in comparison to typical photographic ob
jectives optimized for spectral band from about 450 nm
to about 630 nm.
K. Chrzanowski
167

4. Optimization for input glass optics of IITs. As it was ex
plained in Sect. 3.3 optical objective of NVD must cor
rect influence of input glass optics of IIT on image cre
ated at photocathode of this IIT.
5. High resolution. It is typically expected that resolution
of objective for NVDs should be about two times better
that resolution of IITs. In case of IITs of resolution equal
to 64 lp/mm it means that resolution of optical objective
should be at least 128 lp/mm. For comparison it should
be noted that photographic objectives of resolution hi
gher than about 100 lp/mm are considered as very good.
6. Optimization for cooperation with large sensors. Optical
objectives for NVDs should generate sharp image on en
tire area of photocathode of IITs. Typical diameter of
IITs is 18 mm but tubes of 25 mm photocathode diame
ter or bigger are sometimes used. It should be noted that
sensors used in most of electronic CCTV cameras are
much smaller. Sensors of size equal to 4.83.6 mm are
used in great majority of CCTV cameras.
7. Low mass. The latter requirement is less important in
case of NVDs used as observation systems or as sights in
mechanical vehicles but is of crucial importance in case
portable goggles or sights carried by humans.
It is extremely difficult to fulfil all these requirements at
the same time and design of high performance objectives for
NVDs is considered as one of most challenging tasks for
optical designers.
4.1.2. Ocular
There are five main requirements on oculars for modern
NVDs:
1. Large exit pupil. It is typically expected 14 mm pupil at
25 mm distance from input plane.
2. Wide field of view. It is typically required that the ocu
lars should project high resolution image of the tube
screen in wide field close to 40.
3. Optimization for curvature of output optics of IITs. Ma
jority of tubes used on modern NVDs use curved output
optics and oculars must be optimized for such an input
curvature in order to project a sharp image generated at
screen of such tubes.
4. High resolution. The latter parameter of optical oculars
should be significantly higher (typically two times be
tter) than resolution of image intensifier tubes. In
case of oculars for NVDs built using highres tubes it is
required that ocular resolution is higher that about 120
lp/mm.
5. High transmittance. In order to minimize light losses on
oculars transmittance higher than about 0.95 is expected.
The requirement is typically fulfilled by reducing num
ber of lenses (replacement of spherical lenses by asphe
rical lenses) and improved antireflection coatings.
6. Low mass. The latter requirements is achieved by using
thin glass lenses or by using ultra light plastic lenses.
Fulfilling all these requirements on oculars for NVDs at
the same time is difficult and design of such oculars is typi
cally a compromise between these requirements.
168
4.1.3. Beam splitter

Beam splitter is used in monocular NVG to split and project
image from screen of a single IIT into two optical channels
optimized for two human eyes [(Fig. 6(b)]. Splitting image
generated by IIT into two images automatically reduces per
ceived brightness gain of NVG by a factor of two. Practi
cally, there are some additional losses due to limited trans
mittance of the beam splitter. Therefore, high transmittance
of this optical module at a level over 90% is critical to assure
acceptable brightness gain of such goggles. Next, beam
splitter should be designed in a way that minimizes the
vignetting of propagated optical beam. If the latter condition
is not fulfilled the centre of FOV is significantly brighter
than peripheral parts of image seen by human observer.
4.1.4. Optical filter
Optical filter is an optional module used to modify spectral
sensitivity of NVD in comparison to spectral sensitivity of
IIT.
Spectral sensitivity of majority of NVDs is the same as
spectral sensitivity of IIT used in NVD. Such situation oc
curs because most NVD are built by using optical modules
of flat spectral transmittance in VIS/NIR range. Practically,
it means that typical NVDs are sensitive in spectral band
from about 400 nm to about 850 nm, if Gen 2 tube is used,
or in spectral band from about 500 nm to about 900nm if
Gen 3 tube is used.
This wide spectral sensitivity band is not acceptable in
two cases:
1. NVDs are to be used to help pilots to navigate helicop
ters/aircraft at night conditions;
2. NVDs are to be protected against common lasers.
Reasons for modification of spectral sensitivity of aviator
NVGs are explained in Sect. 7.2. Aviator NVG are typically
equipped with a long bandpass filter that blocks light of wa
velength below about 630 nm (Class A filters), below about
650 nm (Class B filters). Class C filters are more sophisti
cated as these filters are a combination of a long bandpass fil
ter (blocking below about 680 nm) and a bandpass filter (par
tial transmission in band from about 530 nm to 570 nm).
Protecting NVDs against lasers is needed to eliminate
possibility that pilots, drivers, snipers or other peoples using
NVDs are blinded using commonly available lasers; or even
that NVDs are destroyed. Such a protection is achieved
using special optical filters that are expected to attenuate
incoming laser radiation. Precision requirements on such
protecting filters in military NVDs are not known publicly.
Such data are considered as a secret in USA and dissemina
tion of any information related to this technology is prohib
ited [23]. However, it can be logically expected that these
protecting filters are manufactured as a combination of an
interference filter with a substrate of absorptance depending
on power of incoming light.
The interference filter should have several absorption
peaks optimized for wavelengths of typical lasers in visi
ble/NIR range. The substrate should offer broadband ab
sorption covering spectral band of NVD and attenuation of
2013 SEP, Warsaw
this substrate should be at least proportional to power of

incoming light. Short temporal inertia at a level of several
milliseconds is also a critical condition. It is possible that
protecting filters in hightech military NVDs are modified
filters used for protecting eyes of welders.
Protecting filters to be used in NVDs for civilian market
are much simple. Here requirements are generally limited to
attenuation of green and blue lasers. Practically, these are
requirements for long pass filter blocking visible light with
exception of red light. Such protecting filters are commer
cially available and can be easily added to typical NVGs
[24]. However, practically such commercialgrade protect
ing filters offer very limited protection because NVD
equipped with a such filter is still vulnerable to red lasers
and lasers operating in NIR range.
4.2. Influence on NVD performance

NVD is a system built from the main modules: optical
objective, IIT, and optical ocular. Both photometric parame
ters [brightness gain (BG), SNR] and image quality parame
ters [modulation transfer function (MTF), resolution] of
complete NVD depend on parameters of all these modules.
Brightness gain (BG) of complete night vision devices
can calculated as
LG t ob t oc
(1)
BG =
4F 2 +1
where LG is the luminance gain of IIT, tob is the objective
transmittance, and toc is the objective transmittance.
A similar equation can be used to calculate SNR of com
plete NVD
SNRNVD = SNRIIT
t ob
4F 2 +1
(2)
where SNRNVD is the signal to noise ratio of the complete

NVD, and SNRIIT is the signal to noise ratio of the tube.
Both equations show importance of Fnumber of optical
objective for photometric parameters of NVD. By using
such ultra bright objectives both BG and effective SNR
complete of NVD can be improved several times. If typical
F = 1.2 objective is replaced for brighter F = 0.9 objective
than both parameters can be improved 1.6 times. If ultra
bright F = 0.65 objective is used, then improvement equal to
2.5 times can be achieved. Practical improvements are
lower than values stated above but this data clearly show a
method to design ultra sensitive NVDs. The method is prac
tically used by some manufacturers, mostly to design night
vision goggles for aviators and special operation teams [25].
There are, however, several drawbacks of such ultra bright
objectives: higher price, higher dimensions, higher mass.
These drawbacks limit present applications of these objec
tives but it should be expected that such objectives become
more popular in near future.
MTF is a parameter that can be used to describe the abil
ity to generate perfect images by both complete NVD, IIT,
optical objective, and optical ocular.
Relationship between MTF of a complete NVD and

MTF of modules of NVD can be presented in the following
form
MFTNVD (n) = MFTob (n) MFTIIT (n) MFToc (n),
(3)
where MTFNVD is the MTF of complete NVD, MTFob is

the MTF of optical objective, MTFNVD is the MTF of IIT,
MTFoc is the MTF of optical ocular, and n is the spatial fre
quency.
Practical experience of the author shows that MTF of
both NVD and their modules can be usually well approxi
mated by function below
n
MFT(n) = 1
n max
(4)
where nmax is the spatial frequency when MTF is zero.

Next, resolution of both NVD and IIT can be approxi
mately determined as spatial frequency when MTF is about
0.04. Further on, resolution of optical modules can be ap
proximately determined as spatial frequency when MTF is
about 0.02. Higher value of required MTF needed to resolve
resolution target in case of NVD and IIT is due to presence
of significant noise in images generated by these two
devices.
Finally, resolution of NVD can be presented in both
lp/mm units or lp/mrad units. If the latter unit is to be used
then resolution in lp/mm units should be multiplied by
objective focal length presented in meter unit.
Table 6 and Figs. 2223 present results of an analysis of
two NVDs built using the same IIT but using two different
optical modules: a) highres optics and b) typical optics. The
results show clearly that by using high quality optics' im
provements in resolution of NVD at a level of at least 20%
can be achieved.
Table 6. Resolution of two NVDs built using the same IIT but dif
ferent optical modules.
Resolution
Hires NVD
IIT [lp/mm]
64
Objective [lp/mm]
128
Ocular [lp/mm]
170
NVD [lp/mm]
48
NVD [lp/mrad]
1.2
Typical NVD
64
86
105
40
1
Comments
Assumed
input data
Calculated
parameters
of NVD
Resolution of NVD in lp/mrad units showed in Table 6

is calculated for a case of an objective of 25 mm focal
length. This is a typical objective in night vision goggles of
40 FOV built using 18 mm tube. The obtained resolution of
NVD built using high quality optics (1.2 lp/mrad) corre
sponds quite well with resolution value (1.3 lp/mrad) of real
goggles presented in a catalogue by one of big manufactu
rers of NVDs [4].
The results presented in this section show clearly that it
is possible to design a high performance NVD only if all its
modules perform very well. It is possible to find on market
K. Chrzanowski
169
Fig. 22. MTF of NVD built using IIT of 64 lp/mm resolution, and
hires optical modules (objective of 128 lp/mm resolution, and ocu
lar of 170 lp/mm resolution).
Fig. 23. MTF of NVD built using IIT of 64 lp/mm resolution, and
typical resolution optical modules (objective of 86 lp/mm resolu
tion, and ocular of 105 lp/mm resolution).
NVDs built using expensive, high performance tubes and

rather poor quality optical modules. The effect is no more
than Medium performance of such a NVD. Inverse situation
is also possible. Medium performance tubes cooperating
with high performance optical modules can produce quite
good NVD, as well.
To summarize, it is recommended to treat very seriously
night vision optics, measure most critical parameters of opti
cal modules (onaxis MTF, off axis MTF, transmittance, F
number, FOV, vignetting). This measurement data com
bined with parameters of IIT enables to carry out design opti
mization NVD and development of devices with high ratio of
image quality to cost.
Market of night vision devices can be divided into three

major sectors: military applications, law enforcement appli
cations, and commercial applications (hunting, recreation,
astronomy, science).
Theoretically, the border between the first two sectors
(military and law enforcement) and the third sector (com
mercial applications) has been relatively precisely deter
mined by regulations on export control of NVDs issued by
U.S. or EU authorities. Practically, this border can be some
times blurred due to several reasons:
1. Some technically backwards military still use old type
NVDs of parameters below level of good commercial
night vision devices;
2. Some dealers of NVDs active on commercial market ad
vertise their products as military grade goods in situation
when parameters of their devices are below the level
mentioned in export control laws;
3. Law enforcement agencies often purchase on commer
cial market in order to save funds as military grade
NVDs are more expensive;
4. There has been little progress in night vision metrology
within last several decades in spite of big progress in
night vision technology. It is quite common to find on
the world market two NVDs (or two IITs) of the same
data sheet parameters but of a totally different perfor
mance. Inverse situation is possible, too. Next, it is quite
common that test systems used by different manufac
tures of NVDs or IITs generate significantly different
measurement results due to a series of reasons [53].
To summarize, different claims about ultra performance
of so called military grade NVDs should be treated with
caution unless the claims are verified by practical tests.
6. NVD for airborne applications

6.1. Types of aviator NVG
Night vision goggles for airborne applications can be classi
fied as two basic types:
1. Direct view image NVG (Type I);
2. Projected image NVG (Type II).
Type I goggles are NVG that display the intensified
image on a phosphor screen in the users direct line of sight
(Fig. 24). These goggles are often called aviator's night
5. Targeted market
The origin of night vision technology is clearly military.
Several decades ago night vision devices were manufac
tured almost exclusively for military market. Nowadays the
situation on the market has changed significantly. Military
customers are still very important for the manufacturers
especially that they usually purchase most expensive units
but other markets have become important, too. It can be
even said that there are some manufacturers of NVDs that
specialize in nonmilitary markets.
170
Fig. 24. Type I (direct view) night vision goggles with lookaround
vision into the cockpit: (a) diagram and (b) photo (after Ref. 3).
2013 SEP, Warsaw
Fig. 25. Type II (projected image) night vision goggles with look
through outside viewing and optional lookaround vision into
the cockpit: (a) diagram, and (b) photo (after Ref. 26).
vision imaging system (ANVIS). Type I look very much

likes like a pair of binoculars attached to a helmet. These
goggles represent a majority of airborne NVGs, particularly
in case of helicopter pilots.
Type II goggles are NVG that projects the intensified
image through medium in the users line of sight (Fig. 25).
This configuration allows for simultaneous viewing of the
intensified image and visual cues such as headup display
(HUD a semi transparent display that generates data)
symbology [8]. In the past Type II goggles were used exclu
sively for pilots of fixedwing aircraft. Nowadays they are
used also by a growing number of helicopter pilots.
The design of Type I goggles is a cost optimized solu
tion that is characterized by one most important disadvan
tage: a pilot must constantly change its attention from image
generated by the goggles to images from other imaging sys
tems (FLIRs etc) or just to look on helicopter controls.
Therefore, during last several decades a dozen or more Type
II goggles were proposed to eliminate the earlier mentioned
disadvantage of Type I goggles.
The Cats Eyes goggles were the first relatively widely
used Type II goggles. The goggles allowed for the combina
tion of both a direct visual and an intensified image to be
presented to the pilot's eyes. The two images are combined
in a 1:1 relationship and complement each other. The head
up display and helicopter control panel is seen through the
direct visual path. Additionally, the direct vision path thro
ugh the optical combiner arrangement makes easy monitor
ing of cockpit displays. The direct vision path also removes
problems normally associated with light to dark transitions
as the intensified image becomes progressively more notice
able as the direct visual image fades. The system is compact
and rugged and the restrictions on head mobility imposed by
the depth of conventional NVG systems is avoided [27].
However, the folding of the Cats Eyes optical system results
in reducing FOV to a circular 30 FOV with some clipping
of the image in the lower right and lower left. This creates a
situation when pilot is looking like via narrow tunnel. In
addition the beam splitter used to combine two images
reduces also significantly luminance gain of the night vision
goggles.
TopOwl helmet is the latest type of Type II imaging

system developed for airborne applications by Thales. The
Helmet Mounted Sight and Display is based on a unique
concept incorporating a night vision system with a 100%
overlapped projection of a binocular image on the visor.
TopOwl projects the night scene and associated symbology
onto two circular reflective surfaces with a fully overlapped,
40degree, binocular FOV [28]. With training, the pilots
also can look through the visor to read moving maps and
other information on headdown displays. The systems true
binocular presentation and the spreading of the image inten
sification tubes at a distance wider than the spacing of the
pilots eyes help to give a 3D feel to the 2D night vision
imagery. TopOwl has the ability to project thermal and
night vision imagery to the visor, so that pilots and copilots
can select either view. Standard symbology is used to dis
play flight and weapon management data, helping to reduce
crew workload. TopOwl's head position sensor, used during
target designation and weapon firing sequences, enhances
operational effectiveness. This system is also characterized
by low weight and perfect balance of the ergonomically
designed helmet sight.
6.2. Compatibility to aviation regulations

IITs used in night vision NVGs are sensitive to radiation in
both visible and near infrared range. Gen 2 tubes are usually
sensitive from about 400 nm to about 880 nm when Gen 3
tubes are sensitive from about 500 nm to about 920 nm. This
means that NVGs are potentially sensitive to visible light
emitted by helicopter/aircraft control panel.
The light emitted by control panel reflected by glass
window can produce severe veiling glare that can obscure
the overall image for goggles looking through such helicop
ter/aircraft window. If light level from the control panel is
high then automatic gain control mechanism built in NVG
reduces luminance gain of the goggles. The final effect is
that the pilot can see only image of the control panel ref
lected by the window image of real targets outside the
widow becomes dark. In order to prevent such a situation
NVGs for airborne operations must be compatible with
cockpit lights of helicopters/aircraft.
To achieve compatibility of the cockpit lighting with the
night vision goggles, the cockpit lighting should have a
spectral radiance with little or no overlap into the spectral
sensitivity band of NVG. In other words the cockpit lights
should emit very little light in the spectral region in which
the goggles are sensitive. At the same time the lights must
also be visible to the unaided eye in order to enable the pilot
to easily view the cockpit. Such a situation can be achieved
when both cockpit lights and NVGs are optimized for air
borne applications as shown in Fig. 26 [29].
Compatibility of NVG with cockpit lighting can be achi
eved in different ways. NVGs compatible with aircraft/heli
copter cockpit lighting are divided into three classes: Class
A, Class B, and Class C [29]. Spectral sensitivity curves of
these three classes of NVG are shown in Fig. 27.
K. Chrzanowski
171
Fig. 26. General concept of compatibility of night vision goggles

with cockpit lights.
Class A NVG are built like typical goggles with addi

tional so called 625 minusblue filter. Such goggles are not
compatible with red lights because of the overlap between
the spectrum of red light and the sensitivity of Class A gog
gles. Other colour lights as blue, green, orange are accept
able. Most NVGs for airborne applications are Class A
goggles.
Class B goggles are built using 665 minusblue filter.
Such goggles are compatible with all colours of visible
light: blue, green, orange and red and typical cockpit lights
can be used. Such a situation was achieved because there is
practically no overlap between spectral sensitivity curve of
filtered night vision goggles and visible band. However,
sensitivity (and related parameters like luminance gain,
SNR) of Class B goggles is lower than sensitivity of Class A
goggles because of more narrow spectral sensitivity band.
Class C goggles are built using special filters that sup
press blue, orange and red light but transmit partially green
light. Class C was introduced to allow the night vision gog
gles to see HUD symbology presented usually in green col
our. If Class A goggles or Class B goggles are used then this
symbology cannot be seen.
7. Parameters of night vision devices

NVDs generate images that can be seen by humans and it is
possible to evaluate such devices using human sight. How
ever, it is surprisingly difficult even for an expert to pre
cisely evaluate NVDs only by looking on images of typical
scenery. Measurement of a series of parameters is needed in
order to accurately evaluate quality and possible perfor
mance of these devices.
Different sets of parameters of NVDs are used in differ
ent documents: data sheets presented in websites of manu
facturers of NVDs [2,4,5,6,25,30,31], military standards
[3244], tender technical specifications [4547], export
control regulations [48,49], in aviation technical recommen
dations [50], or in educational sections of manufacturers'
websites of NVDs' testing equipment [51,52]. On the basis
of analysis of these literature sources, parameters of NVDs
can be divided into twelve groups: parameters of IITs, para
meters of optical modules, image quality parameters, photo
metric parameters, geometrical parameters, blemishes, bin
172
Fig. 27. Relative spectral sensitivity of night vision goggles compat

ible to lighting in aircraft/helicopter cockpits (after Ref. 29).
ocular parameters, operational defects, environmental para

meters, mechanical parameters, electrical parameters, and
function checks. Detailed list of parameters of night vision
devices is shown in Table 7.
It is beyond scope of this paper to present here a review
of parameters of night vision devices. The readers should be
however warned that characterization and testing NVDs is a
difficult task, probably more difficult than testing thermal
imagers. Present situation in night vision metrology is not
optimistic in spite of significant technological advances in
night vision technology during last several decades [53].
The readers interested in characterization and testing NVDs/
IITs are encouraged to read literature sources mentioned
above in this section.
8. Technology trends
Forecasting technology trends in night vision technology is
risky. However, on the basis of analysis of data available
from manufacturers of NVDs and manufacturers of IITs,
seven separate development trends can be determined:
1. Digital NVDs;
2. Enhanced (dual sensor) NVDs;
3. Active NVDs;
4. Ultra sensitive NVDs;
5. NVD of enlarged field of view;
6. Ultralight NVDs;
7. Low cost NVDs.
The first three trends represent basically new types of
night vision systems. The latter four trends are generally
improvements of classical NVDs.
8.1. Digital NVDs

Classical night vision devices built by using image intensi
fier tubes are sensitive in similar spectral range as surveil
lance visible/NIR cameras built by using solid state silicon
sensors. At the same time visible/NIR cameras are elec
2013 SEP, Warsaw
Table 7. Parameters for characterization of night vision devices.

No
Group of parameters
List of parameters
Parameters of image
intensifier tubes
resolution, MTF, SNR, blemishes, image alignment, halo, etc these parameters are to be evaluated us
ing tube certificates
Parameters
of optical modules
resolution, MTF, FOV, transmission these parameters are to be evaluated using certificates
Image quality
parameters
resolution, MRC (minimal resolvable contrast), range of focus, eyepiece diopter range
Photometric parameters
brightness gain, saturation level
Geometrical parameters
field of view, magnification, distortion
Blemishes
dark spots, white spots, fixed pattern noise, chicken wire, output brightness variation
Binocular parameters
collimation error, gain balance, magnification balance, distortion balance image rotation balance,
interpupillary distance
Operational defects
shading, edge glow, flashing/flickering/intermittent operation, and emission points
Environmental
parameters
extreme temperatures, temperature shock, altitude, humidity, vibration, shock, sand and dust, fungus,
immersion, salt fog, explosive conditions, transportability, electromagnetic interference (EMI)
10 Mechanical parameters
weight, size, centre of gravity
11 Electrical parameters
power voltage, battery type, battery operational time
12 Function checks
electrical switch, objective torque, Eyepiece torque, LED indicator, high light cut off, low voltage indi
cator, interchangeability, flipup/flipdown (aviator goggles), automatic breakaway (aviator goggles)
Fig. 28. Digital NVDs: (a) Pulsar Hindsight N550 digital sight, (b) Ghost Hunter 550 monocular (after Ref. 54), and (c) MonieD goggles
(after Ref. 55).
tronic imaging devices that inherently possess some impor

tant advantages as image processing, storing, automatic fo
cusing, automatic target recognition, ability for electronic
communication, etc. Next, some of visible/NIR cameras
offer sensitivity similar to classical NVDs. Therefore, digi
tal night vision devices built by using high sensitive silicon
imaging sensors and miniature electronic displays should be
theoretically a much better alternative to classical night
vision device based on optical image intensification.
Digital NVDs are already offered on commercial market
as sights, monoculars, and goggles [54]. There are manufac
turers of classical NVDs for military sector that offer also
digital monocular goggles [55]. However, so far a share of
digital NVDs on commercial market is small (it can be esti
mated below 1%); a share of digital NVDs in military mar
ket is even smaller. Digital NVGs are treated on both mar
kets as an interesting novelty that must be verified both
technically and economically.
Digital NVDs looks externally the same as their classi
cal analogues (Fig. 28). Such a situation occurs because
external modules (objective, ocular, mechanical case) are

the same for both types of NVDs. However, practically
there are big differences between internal modules used by
these two types of NVDs. As it is shown in Fig. 29 new
modules (image sensor, signal processing, and display) are
used in digital devices to replace a single module (IIT) used
in classical NVDs.
Image sensor (typically integrated with signal process
ing electronics) and miniaturized display are two most
important modules of digital NVDs.
K. Chrzanowski
Fig. 29. Block diagram of digital NVDs.
173
8.1.1. Image sensors

There are at least seven technologies of silicon imaging sen
sors that compete in digital night vision:
1. Intensified CCD (ICCD);
2. Intensified CMOS (ICMOS);
3. Electron multiplied CCD (EM CCD);
4. Electron bombarded CMOS (EBCMOS);
5. Microchannel plate complementary metaloxide semi
conductor (MCPCMOS);
6. Scientific CCD (sCCD);
7. Scientific CMOS (sCMOS).
ICCD/ICMOS module is basically an IIT coupled to
classical CCD (or CMOS) imaging sensor by fibre optics
with additional some electronics for processing signal gene
rated by CCD/CMOS sensor (Fig. 30). Digital NVDs based
on ICCD sensor are the oldest type of these digital devices.
EMCCD is a technology that uses a solid state electron
multiplying (EM) register at the end of the normal serial
register [57]. The EM register allows weak signals to be
multiplied before any readout noise is added by the output
amplifier. The EM register has several hundred stages that
use higher than normal clock voltages. As charge is trans
ferred through each stage, the phenomenon of impact ion
ization is utilized to produce secondary electrons, and hence
the name of EM gain. In this way EMCCD sensors are capa
ble of detecting single photon events. In this way EMCCD
eliminates important shortcoming of conventional highper
formance CCD cameras that very low signal levels typically
fall beneath the read noise floor of the sensor.
EMCCD technology originated at beginning of 2000s.
Now it is a fully matured technology offered by a series of
companies. The problem is however that EMCCD sensors
are typically cooled by using thermoelectric cooler (or other
types of coolers) in order to reduce its temperature and con
sequently its noise. The latter feature makes EMCCD cam
eras rather bulky. Due to big size, mass and high power con
sumption EMCCD sensors have no chance to find mass
applications in portable digital NVDs but can be potentially
used in digital NVDs as observation systems in mechanical
vehicles.
Electron bombarded active pixel sensor (EBAPS sensor)
is an ultra sensitive imaging sensor based on a IIIV semi
conductor photocathode in proximityfocus with a highres
Fig. 30. ICCD module: (a) block diagram and (b) photo (after
Ref. 56).
174
olution, backsidethinned, CMOS chip anode [58]. The

electrons emitted by the photocathode are directly injected
in the electron bombarded mode into the CMOS anode,
where electrons are collected, amplified and readout to pro
duce digital video directly out of the sensor. This technol
ogy enables a design of a compact, lightweight and low
power sensor offering high quality images even at a very
low illumination level. Low power consumption of EBAPS
sensors is a significant advantage over EMCCD technology.
Due to low mass and power consumption EBAPS tech
nology enables design of portable NVDs.
EBAPS is a patented technology and EBAPS sensors
are manufactured by only one manufacturer Intevac Corp.
This company has already started production of both digital
night vision monoculars and digital binocular NVGs [58].
Microchannel plate complementary metaloxide semi
conductor (MCPCMOS) is a technology developed similar
to EBAPS technology. MCPCMOS sensor uses modified
CMOS chip directly into the vacuum envelope of a proxim
ity focused image tube like typical EBAPS sensor. The pri
mary difference is that the MCPCMOS sensor contains a
microchannel plate and, therefore, can offer higher lumi
nous gain capability [59]. This technology has been devel
oped by ITT Night Vision a big manufacturer of classical
IITs that clearly shows interest of this company in digital
night vision.
sCCD/sCMOS sensors are so called scientific grade
CCD (or CMOS) imaging sensors of significantly reduced
defects and blemishes, reduced several times internal noise,
improved image resolution and expanded dynamic range in
comparison to typical CCD/CMOS sensors used in com
mercial video cameras. The price of scientific grade CCD/
CMOS sensors can be even over 100 times higher than the
price of typical sensors.
sCCD/sCMOS sensors are often used in high end imag
ing systems used in science and industry. Due to recent
improvements in sensitivity sCCD/sCMOS both sCCD and
sCMOS sensors are considered as serious competitors of the
earlier mentioned sensor technologies in digital night vi
sion. sCMOS sensors have already found mass applications
in highend colour surveillance cameras for night applica
tions. These small, rugged, lowpower cameras can generate
Fig. 31. Digital sCMOS sensor called Lynx offered by Photonis

(after Ref. 61).
2013 SEP, Warsaw
1280960 highdefinition (HD) image, offer ultrahigh

25000:1 dynamic range and less than 2e (rms) read noise
[60]. One of big manufacturers of classical IITs is offering
also sCMOS sensors of 12801024 resolution for digital
night vision designed to provide excellent performance in
low light level conditions (read noise below 4e, power
consumption under 200mW) [61].
Simpler imaging modules built by using the technolo
gies mentioned above generate output image in typical ana
logue television formats: PAL or NTCS (720576 at 25
frame rate for PAL format or 640480 at 30 frame rate for
NTSC format). Additional analogue/digital module is used
to convert analogue images to digital form that is later sent
to a display module. More advanced imaging modules gen
erate output image using digital image formats: DVI, USB
2.0 or Camera Link. Images of resolution over 12801024
at 25 frame rate can generated and sent directly to the
display module.
8.1.2. Miniaturized displays

There are two main technologies of miniaturized displays used
in digital NVDs: thin film transistor liquid crystal display
(TFTLCD) and organic light emitting diode displays
(OLED).
TFTLCD displays or more accurately active matrix
TFT LCD (AM TFT LCD) are electronic devices that gen
erate image on array of LCD pixels using for control active
matrix of thin film transistors. Voltage dependent optical
polarisation of LCD crystal is used as a basic phenomenon
that enables image generation on LCD array.
OLED displays use of electroluminiscence phenomenon
of some organic compounds. A film of organic compound
that emits light in response to an electric current is used to
generate image.
There are two main families of OLEDs: those based on
small molecules and those employing polymers [83]. Add
ing mobile ions to an OLED creates a lightemitting electro
chemical cell (LEC), which has a slightly different mode of
operation. Next, OLED displays can use either passive
matrix (PMOLED) or activematrix addressing schemes.
Activematrix OLEDs (AMOLED) require a thinfilm tran
sistor backplane to switch each individual pixel on or off,
but allow for higher resolution and larger display sizes.
Both TFTLCD and OLED technologies are now used to
manufacture big numbers of minidisplays for commercial
and military applications. However, it is typically considered
OLED displays are a better choice for both these application
areas.
Commercial OLED displays are thinner, lighter and
more flexible. Next, the OLED displays offer higher view
ing angle, brighter images, more vivid colours, higher con
trast (both dynamic range and static), faster colour switch
ing, reduced ghosting or blurring effects, short response
time (~ 0.01ms,) and consume less power in comparison to
LCD. However, OLED displays are characterized by more
limited life time than LED displays [62]. Ability of OLED
display to work at low temperatures without heaters, elimi
nation of backlights, display flexibility are considered as a

big advantage of OLED displays in military applications
[63].
8.1.3. Technological challenges

The presented earlier long list of technologies used in digital
night vision looks impressive but practically there are seri
ous problems to design digital NVDs of sensitivity, image
resolution, compactness, power consumption and cost simi
lar to parameters offered by classical NVDs.
The requirement on similar sensitivity is not a big tech
nical challenge. Cameras manufactured using all technolo
gies mentioned earlier generate images of minimal degrada
tion of image quality at illumination as low as 10 mlx (Quar
ter Moon) [64]. Next, CCD/ICMOS/ EMCCD/EBAPS tech
nologies offer sensitivity similar to typical NVDs and po
tentially enable surveillance at illumination as low as 1 mlx.
The biggest technical challenge for designers of digital
NVDs is to achieve image resolution similar to resolution of
classical NVDs. It should be remembered that classical IITs
generate images of equivalent resolution of about 2 500 000
pixels (IIT of 18 mm photocathode, and modest resolution
equal to 50 lp/mm) at the frame rate at about 25 frame per
second. In order to enable design of digital NVD offering
image of resolution equivalent to classical NVD two crucial
modules are needed: high resolution imaging sensor and
miniaturized high resolution display.
There are available imaging sensors of resolution
exceeding 2 500 000 pixels' limit. However, such sensors
are very expensive and not suitable to be used in digital
NVDs for mass applications. Imaging sensors of pixel num
ber 12801024 (about two times less than earlier mentioned
number of equivalent pixels of an image in typical NVDs)
can be considered as current resolution limit of imaging sen
sors that can be used in digital night vision.
Ten years ago miniaturized displays were considered as
the main obstacle to design digital NVDs even of modest
resolution of typical analogue TV images. Nowadays, situa
tion on market of miniaturized OLED displays looks much
better, even impressive due to recent dramatic improve
ments in resolution of miniaturized OLED displays [65].
Bicolour and tricolour OLED displays of resolution up to 5
bilion dots are now commercially available.
Requirement on compactness represents another techni
cal challenge for designers of digital NVDs. It should be
remembered that weight of classical monocular NVDs can
be as low as 250 g. Size of imaging sensors was never a big
limitation for digital night vision. Situation with size of dis
plays was inverse. Display size was one of factors that made
impossible to design a portable NVD one decade ago. How
ever, last decade is a period of big achievements in minia
turization of both TFTLCD and OLED displays. Nowa
days, displays of sizes smaller than screens of IITs are man
ufactured and are commercially available (Fig. 32) [65].
Power consumption of imaging sensors and displays in
digital NVD is a difficult problem, not fully solved, so far.
Power consumption of typical NVD is often below 0.1W.
K. Chrzanowski
175
Fig. 32. Photo of a microdisplay of diagonal size equal to 0.38

inches (after Ref. 65).
Power consumption of digital NVDs is typically in a range

from about 500 mW to about 1200 mW. It means that cur
rently power consumption of classical NVDs built by using
IITs is at least five times smaller than power consumption of
digital NVDs. The problem of power consumption of digital
NVDs is only partially neutralized by big progress in tech
nology of miniaturized batteries in the last decade.
Cost is another, probably, the biggest problem for digital
night vision. At present, it is not possible to replace a single
IIT module by an equivalent set of three modules (imaging
sensor, signal processing electronics, miniaturized display)
and keep the same manufacturing price.
To summarize, digital NVDs are an interesting novelty
on market that present several potential advantages over
classical NVDs. However, these potential advantages of
digital night vision are not so far fully materialized because
of both technical and economical challenges. Practical tests
often show clear superiority of classical NVGs from the
point of view of sensitivity, image resolution, compactness,
and power consumption. There are rumours that users of
digital goggles loose more quickly orientation or balance
when wearing such goggles. At the same time manufactur
ing costs of high performance digital NVDs are higher than
in case of equivalent classical NVDs. Therefore, probability
that digital NVDs will eliminate classical NVDs is low.
However, digital night vision technology is making fast
progress, technical performance of digital NVDs is quickly
improving and it can be expected that market share of digital
NVDs will grow steadily in future.
nition, ability for electronic communication, etc. However,

both classical NVDs and digital NVDs suffer from the same
limitation because both technologies use the same spectral
band. Both devices are practically blind in dense fog, smo
ke, sand storms, or at very low illumination conditions.
Next, these devices are also not effective against classical
camouflage techniques.
Fusion of classical night vision with thermal imaging is
very interesting for both military and commercial users of
NVDs because these new fused devices combine advan
tages of both two technologies:
1. Ability to generate high contrast images of warm targets
of interest at low illumination/poor atmosphere condi
tions (detection task);
2. Ability to generate high resolution image of both targets
and background (identification task).
There is a huge interest in enhanced night vision due to
advantages mentioned earlier. Enhanced NVDs are already
commercially available (Fig. 33). These new devices offer
high contrast images like images shown in Fig. 34. Human
users can easily see high contrast warm features (typically
bright orange or red colour) over typical green background
in a single image and can easily recognize hidden targets.
There is a bright future for enhanced night vision tech
nology due to earlier mentioned advantages. However, at
present there are big technical and economical challenges
that this technology must overcome in order to be used in
high numbers.
8.2. Enhanced NVDs

Enhanced NVDs (dual sensor NVDs) are probably the most
important trend in night vision technology.
Digital NVDs presented in the previous section offer
some advantages over classical NVDs due to image pro
cessing, storing, automatic focusing, automatic target recog
176
Fig. 33. Enhanced night vision devices: (a) enhanced monocular

goggles (after Ref. 4) and (b) enhanced binocular goggles (after
Ref. 66).
2013 SEP, Warsaw
Fig. 34. Images of forest scenery at foggy conditions generated by

classical NVGs (left) and ENVG (right) (after Ref 67).
Details of design of enhanced NVDs shown are consid

ered as secret [68]. Monocular goggles designed as a com
bination of two optical channels (digital night vision chan
nel and thermal imager channel) shown in Fig. 35 are one of
possible design solutions.
All technical challenges for digital night vision are still
valid for enhanced night vision because digital NVD is one
of channels of enhanced NVD. Therefore, only additional
technical problems specific for enhanced NVDs will be pre
sented in this section.
Resolution of IR FPA sensors used in thermal imagers
has always been behind resolution of silicon arrays (CCD/
CMOS/ICCD/EMCCD/EBAPS/sCCD/sCMOS) used as
imaging sensors in a visible and near infrared range. IR
FPAs of 1024768 resolution can be considered as current
technology limit for sensors to be used in enhanced night
vision. It is several times below resolution of classical
NVDs but still acceptable due to ability of IR FPA sensors
to produce high contrast images.
Noncooled IR FPA sensors are the only option for en
hanced night vision due to too high power consumption and
price of cooled IR FPAs. Power consumption of modern non
cooled IR FPA of 1024768 resolution can be as low as
about 100 mW [69], but overall power consumption of com
plete thermal module is much higher and limiting power con
sumption below about 400 mW is a real technical challenge.
Next, it is technically possible to use two displays for two

channels in enhanced NVDs but solution with a shared dis
play and optical image splitter (shown in Fig. 35) is preferred
due to effort to limit power consumption.
Design and manufacturing of high resolution, ultra light,
ultra bright (F number about 1) objective capable to cooper
ate with high resolution of 1024768 (17 m pixel) in full
FOV is another difficult problem for designers of enhanced
NVDs. There are on market very few IR objectives that can
fulfil at the same time requirements on wide FOV, high
modulation transfer function (MTF), low Fnumber, high
transmittance and low weight.
The most difficult challenge specific for enhanced night
vision is fusion of images generated by two channels work
ing at two different spectral ranges.
Humans are equipped with an optical system in form of
two eyes sensitive to light in a visible range. Human brain is
doing image fusion from these two optical channels and cre
ates a single two dimensional image (but with some depth
perception) of targets located at distance from a dozen of
centimeters to several kilometers.
Enhanced NVDs are expected to work in the same way
as the human sight. The challenge is that images of the same
target seen by two channels of enhanced NVD are slightly
different and this difference depends on distance to the tar
get. Advanced processing of images from two channels of
enhanced NVD is needed to find optimal overlaying of
images from these two channels. Next, advanced image pro
cessing is needed to determine which parts of thermal image
should be used in the final fused image, which parts of visi
ble image are eliminated, and how to blend thermal image
with a visible image without decreasing sharpness of final
fused image.
To summarize, enhanced night vision combines advan
tages of two classical technologies (image intensification
and thermal imaging) and is potentially very attractive for
both military and commercial users. However, technologi
cal difficulties are higher than in case of digital night vision.
Present prices of ENVGs can be estimated at a level over
2050 times higher than price of classical NVGs. Price of
NVGs manufactured in mass production is expected in the
future to go up to a level about three times higher than price
of classical NVD [70]. This reduced but still high price
combined with reduced sensitivity and image resolution of
ENVGs in comparison to classical NVDs rather eliminate
possibility that ENVGs will dominate night vision market.
However, it can be expected that ENVGs will take a big
share of night vision market in near future due to a several
of advantages of enhanced night vision.
8.3. Active NVDs
Fig. 35. Simplified block diagram of enhanced NVDs.

First NVDs were active systems that required artificial illu

mination in near infrared range (powerful tungsten search
lights). Such a feature is a significant drawback in military
applications and passive NVDs were developed several
decades ago. Therefore, modern classical NVDs are fully
K. Chrzanowski
177
Fig. 36. Block diagram of an active night vision device.
passive surveillance systems using only in exceptional situ

ation small short range illuminators (map reading).
There are, however, on the market some advanced active
NVDs taking advantage from technological improvements
in laser technology. Modern active night vision devices are
built, like old NVDs, from two main modules: active light
source and passive imaging device. However, simple tung
sten/halogen bulb working in direct mode (the light level is
quasi constant) in old active NVDs is replaced by powerful
laser diodes emitting pulsed light in NIR range in modern
active NVDs. Next, the passive imaging device is now
working in gated mode (sensing device is active only at
short time durations) and is synchronized with the active
light source. This means that the receiver is active only at
time intervals when the pulses emitted by the light source
and reflected by the observed targets are supposed to reach
back the imaging device.
In this way the drawbacks of old active NVDs are almost
eliminated. Due to directional character of emitted light and
short time when the new active night vision systems emit
light probability of detection of these systems is much lower
than case of old active night vision systems. Gating mode
used in modern active night vision systems makes them less
vulnerable to fog, rain or in general to atmosphere condi
tions. Higher range of effective observation can be achie
ved. Further on, modern active night vision system can gen
erate sharp images of targets located close to strong light
Fig. 37. Photo of an active night vision device type DM 740 from
Dedal, Russia (after Ref. 71).
178
sources are located in situation when typical passive night

vision devices or typical passive visible/NIR electronic ca
meras generate only blurred images of the light sources and
black background. Finally, some of modern active NVDs
can work as low accuracy range finders (they produce sharp
image only of targets at regulated expected distance).
In spite of earlier mentioned important advantages these
new active NVDs are not popular in comparison to typical
passive NVDs. However, situation can change in the future
because portable, active range gated NVDs are a small off
spring of a much wider active imaging technology.
Laser detection and ranging (LADAR) cameras, time of
flight (ToF) cameras, active range gated night vision cam
eras are three important groups of active imaging technol
ogy that differ in technical solutions but use the same princi
ple as earlier discussed active NVDs. The latter three main
groups of active imaging have found mass applications in
automotive industry, surveillance, 3D modelling, robotics
etc [7274].
It is beyond the scope of this paper to discuss in details
groups of active imaging system but should be noted that
further development of this technology will bring improve
ments in active NVDs.
8.4. Ultra sensitive NVDs

Recent military conflicts have shown that present genera
tion of NVDs still performs rather poorly at very dark night
conditions (overcast starlight) or even at relatively good
illumination conditions (quarter moon) but in areas of low
terrain reflectance (like mountains of Afghanistan). At pres
ent level of night vision technology dark nights can still sig
nificantly reduce military activities. Some elite units equip
ped with ultrahigh sensitivity NVDs can carry out opera
tions even at overcast starlight illumination conditions but
effectiveness of surveillance is limited.
Present ultrasensitive NVDs are built by using Gen 3
Filmless/Gen 3 Thin Film tubes and ultra bright optical
objectives of Fnumber as low as 0.9. It can be expected that
further technological improvements of sensitivity of Gen 3
Filmless tubes or Gen 3+ Thin Film tubes (SNR as high as
35) combined with improvement of high resolution ultra
fast objectives of ultra low Fnumber (as low as 0.65) will
improve sensitivity of NVDs even more.
As it was shown in Fig. 38 there is a big difference in
performance in dark night conditions between a typical
NVG and an ultra sensitive NVG. The latter NVGs enable
safe flights for helicopter pilots when use of the first ones
can lead to accidents.
Ion barrier film is the main factor that limits further
improvement of Gen 3+ Thin Film tubes and it is doubtful if
these tubes can ever achieve earlier mentioned level of SNR
parameter. It is more probable that the best of Gen 3 Fil
mless tubes will offer SNR equal to 35 or higher in near
future.
Significant aberrations, strong vignetting effect, narrow
focus range, big sizes and mass are four main technological
2013 SEP, Warsaw
Fig. 38. Photo of low contrast mountain landscape at moonless night

obtained using two NVG of different design: (a) NVG built by using
typical F1.2 optics and typical IIT of SNR equal to 22 and (b) NVG
built using F0.9 optics and IIT of SNR equal to 32.
challenges in a design of ultra bright objectives needed for

new generation of ultra sensitive NVDs. Design of high res
olution (aberration blur below about 3 micrometers), ultra
bright (F number as low as 0.65) small, compact, light
objectives of wide FOV (over 40) cooperating with big 18
mm sensors is a very high challenge for optical designers.
However, sooner or later these optical objectives will be
developed and increase significantly sensitivity of NVDs.
8.5. NVDs of enlarged FOV

Horizontal field of view of a human eye is about 40 and
vertical field of view is about 30. Due to its ability to rotate,
a human eye can effectively see (detection of movement)
almost in full hemisphere. Therefore, narrow FOV in com
parison to human eye FOV has always been considered as
one of basic limitations of night vision devices.
The efforts to increase FOV of NVD have always been
present in vision technology. Old standard FOV of NVG
used until about the end of 1980s was 30. Nowadays, great
majority of short range NVDs (goggles, monooculars) offer
40 FOV. However, there are also on market goggles of
wider FOV. Night vision goggles Lucie from Thales Ange
nieux having FOV equal to 50 is an example of this trend
(Fig. 39). It is possible that such goggles become popular in
future and 50 will become a new standard FOV.
Fig. 39. Night vision goggle Lucie of enlarged field of view from
Thales Angenieux (after Ref. 75).
It should be noted that FOV and resolution are con

nected. Increasing FOV means to sacrifice resolution and
inverse. Therefore, goggles of enlarged 50 FOV are char
acterized by slightly lower resolution than typical goggles
of field of view about 40. Next, one of tricks of sellers of
low cost NVGs for civilian applications of 30 FOV is to
advertise their high resolution comparable to resolution of
expensive military grade NVGs of 40 FOV without men
tioning difference in FOV. Table 8 presents the relationship
between resolution of IIT, FOV and resolution of NVD and
shows clearly how it is easy to manipulate with resolution of
NVDs if information about FOV is not available.
Table 8. Relationship between resolution of IIT, FOV and theoreti
cal (case of ideal optics) resolution of NVD.
Resolution Resolution of
of IIT
NVD for 30
[lp/mm]
FOV [lp/mrad]
45
1.20
64
1.71
Resolution of
NVD for 40
FOV [lp/mrad]
0.90
1.28
Resolution of
NVD for 50
FOV [lp/mrad]
0.72
1.02
Panoramic NVG represents another version of a trend to

enlarge FOV of NVGs but without sacrificing resolution
because panoramic goggles are built by using several tubes.
Modern panoramic NVGs are typically built by using four
16 mm tubes and offer FOV about 80100 (horizontal) to
about 35 (vertical).
Experimental projects to design panoramic NVG to dou
ble typical horizontal FOV have been carried out for deca
des. A series of panoramic NVGs for pilots of fixedwing
aircraft has been developed. Tests have confirmed big po
tentials of such goggles due to enlarged FOV but also have
showed a series technical problems encountered in practical
use of panoramic NVGs [7678].
After a long period of experimental tests panoramic
NGVs are nowadays used as standard imaging systems in
several fixedwing aircraft [79]. Panoramic NVGs for avia
tors are commercially available, too [80].
Panoramic NVGs were developed and are mostly used by
aviators. However, it should be noted that panoramic NVGs
for land applications are recently also offered (Fig. 40). It
means that in near future panoramic NVGs can be used in
much higher numbers than they were used in past.
Fig. 40. AN/AVS10 panoramic night vision goggle offered by L3

Warrier Systems (after Ref. 81).
K. Chrzanowski
179
8.6. Ultralight NVDs

Fatigue of soldiers or other users of NVDs wearing heavy
goggles attached to helmets or masks was always consid
ered as a drawback of night vision technology. An ideal
NVD should be of mass comparable to human eyes. Ultra
light NVG or monoculars built by using new small 16 mm
tubes are a potential solution to this limitation. There are
already on market night vision monoculars as light as 250 g
of a very ergonomic design (Fig. 3) [31]. The same can be
said about size and weight as even modern NVDs are very
big and heavy in comparison to human eye.
market. At the same time these new tubes offer images of

much better quality than typical Gen 1 tubes. It is reported
that luminous sensitivity of CORE tubes can be as high as
350 A/lm and resolution close to 60 lp/mm [82]. This per
formance level makes CORE tubes a serious rival of low
cost Gen 2 tubes.
9. Conclusions
Projected image NVG (Type II goggles) are devices that

projects the intensified image on a see through medium in
the users line of sight. These goggles allow for simulta
neous viewing of the intensified image, control panel, HUD
symbology and images from other electronic imaging sys
tems like thermal imagers etc. Nowadays Type II goggles
are part of highly sophisticated systems that combine mod
ules of NVG, projection optics, image combining optics,
helmet, position sensing electronic system, and image pro
cessing electronics. TopOwl system developed by Thales
(Fig. 25) can be treated as a prime achievement in this trend
of night vision technology. It can be expected that simpli
fied, lower cost projected image NVDs will be offered on
market in near future.
Night vision based on technology of image intensifier tubes

is the oldest electrooptical surveillance technology. How
ever, in spite of strong competition from thermal imagers,
visible/NIR cameras and digital night vision, this old mature
technology is still in a phase of growth. There are no signs
of possible demise of classical optical NVDs in near future.
Night vision is a fully matured technology that has fo
und mass applications in both military, security and defence
sectors. NVDs are offered on international market in form
of a long series of devices of different design configuration,
type of image intensifier tube, type of night vision optics,
and performance. Proper understanding and evaluation of
NVDs is a complicated task as many details are to be taken
into account.
This review of modern night vision technology can help
readers to understand sophisticated situation on interna
tional night vision market. However, reading of literature on
characterization and testing of night vision devices, and
analysis of dynamic situation in trends of night vision tech
nology is recommended as a supplement to this paper.
8.8. Low cost NVDs
References
8.7. Projected image NVDs
Price is a prime factor for NVDs at commercial market. Due

to this factor devices built by using Gen 1 tubes are still
used. High price of Gen 2/Gen 3 tubes is the main obstacle
for wider use of NVDs built by using these more modern
tubes for commercial applications. Therefore, efforts to de
crease costs of manufacturing of Gen 2 tubes are continued
and some Gen 2 tubes of lower luminous sensitivity (below
about 350 A/lm) and resolution (below about 45 lp/mm)
are manufactured for commercial market as non military
goods.
Improved Gen 1 tubes represent another direction in
trend to lower cost of IITs for commercial market. Ceramic
optical ruggedized engine (CORE) tubes represent the latest
novelty on commercial market. Gen 1 tubes are character
ized by three important drawbacks: low sensitivity, medium
resolution, and vulnerability to mechanical shocks. Gen 1
tubes made in CORE technology almost eliminate these
drawbacks. The CORE tubes are manufactured by means of
fusion of metal alloys with ceramic compounds similar to
Gen 2 tubes and unlike classical Gen 1 tubes manufactured
using glass technology. In this way ability to produce of
robust night vision weapon sights for restricted budgets was
achieved because CORE tubes shockproof against mid
dlesize riffle calibers. The latter feature is extremely impor
tant for hunters that form significant group of commercial
180
1. A. Rogalski, History of infrared detectors, OptoElectron.

Rev. 20, 279308 (2012).
2. www.atncorp.com
3. International military catalogue Night vision products, ITT
Night Vision, 2010
4. www.nightvision.com
5. www.vectronix.ch
6. www.optixco.com
7. www.inframet.pl/computer_simulators.htm
8. The Avionics Handbook, edited by Cary R. Spitzer, Boca
Raton, 2001.
9. www.nightvisioncn.com/sdp/625512/4/cp3235901/0/Eval
uating_Night_Vision.html
10. en.wikipedia.org/wiki/Image_intensifier
11. Image Intensifiers catalogue no. TII 0004E02 IP, Hamama
tsu Photonics K.K, 2009
12. www.nightvision.com/products/military/case_studygen3.htm
13. www.photonis.com/en/night_vision
14. aunv.blackice.com.au/cgibin/nightvision/forum?index=dis
cussions&story=omni
15. www.exelisinc.com/News/PressReleases/Pages/OmniVIII.as
px
16. L. Bosch, Tube performance that matters, Proc. SPIE
4128, 6578, (2000).
17. MILI49453 CR, Image intensifier assembly, 18 millimiter
microchannel wafer, MX 10130/UV, 1989
18. www.colutron.com
2013 SEP, Warsaw
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
www.stanfordcomputeroptics.com
Introduction to scientific imaging, Roper Scientific Inc., 2002
www.nighttronic.de/Seite116
www.vectronix.ch/#/en/news_press/current_news/indotarsius
www.sidley.com/db30/cgibin/pubs/Statement_of_Facts.pdf
nightflightconcepts.com/pdf/LaserDefense/LaserArmorLIF
_Datasheet_v4_LowRes.pdf
www.angenieux.com/zoomlenses/index.php?rub=5
www.thalesgroup.com/TopOwlH.aspx
H. L. Task, Night vision devices and characteristics, Ams
trong Laboratory Report ASC, 912961 (1992).
www.aviationtoday.com/av/categories/military/TwoWays
toSee_1159.html
MILSTD3009, Lighting, Aircraft, Night Vision Imaging
System (NVIS) Compatible, 2001.
www.transvaro.com.tr
www.vectronix.ch/#/en/products/handheld_equipment/im
age_intensifiers
MILG49313CR, Goggles night vision AN/PVS7B, 1989
MILA49425(CR), Aviator's night vision imaging system
AN/AVS6, 1989
MILPRF49082E, PERFORMANCE SPECIFICATION
VIEWER, DRIVER'S, NIGHT VISION, AN/VVS2(V, 1999
MILPRF49063E, NIGHT VISION SIGHT, INDIVIDUAL
SERVED WEAPON AN/PVS4, 1999.
MILPRF49065F, NIGHT VISION GOGGLES, AN/PVS
5, 1999.
MIL STD 1858, Image intensifier assemblies, performance
parameters of, 1983.
MILPRF49052G Image intensifier assembly, 18 millimi
ter microchannel wafer, MX9916/UV, 1999.
MILPRF49428 Image intensifier assembly, 18 millimiter
microchannel wafer, MX10160/AVS6, 1995.
MILSTD1858, Image intensifier assemblies, performance
parameters of; 1981
MILI49453 CR, Image intensifier assembly, 18 millimiter
microchannel wafer, MX 10130/UV, 1989
GOST 21815.086GOST 21815.1786 Image intensifier
tubes Measurement methods of optical and photometric pa
rameters, 1987 (in Russian).
Stanag no. 4348, Definition of nominal static range perfor
mance for image intensifier systems, 1988 (annulled in
1996).
STANAG No. 4351, Measurement of the minimum resolv
able contrast (MRC) of image intensifiers,1987 (annulled in
1996).
Tender Enquiry No 2/Tech/2007(1) for Night Vision De
vice, Ministry of Home Affairs Government of India.
Request for Proposal (RFP) No : PC28706/ACSFP/PNVB/
A/ARTY dated 01 Dec 10., GOCinC, Northern Command,
C/o 56 APO, India.
Global tender inquiry no TE No. U.II. 871/201011PROC
IV dated the 02 August, 2010, India.
International Traffic in Arms (ITAR) per title 22, Code of
Federal Regulations (CFR), Parts 120130.
EC Regulation No WE 428/2009: Community regime for the
control of exports, transfer, brokering and transit of dualuse
items.
50. Addendum to GEN09ASAM01, Aviation Safety Action

Message, Updated Night Vision Device (NVD) Mainte
nance, 151130Z OCT 09, 2009.
51. www.hoffmanengineering.com
52. www.inframet.com
53. K. Chrzanowski, Present status of night vision metrology,
OPTRO 5th Int. Symp. on Optronics in Defence and Security,
Paris, 2012.
54. www.nightvisionstore.com
55. www.angenieux.com/zoomlenses/index.php?txt=38
56. www.armedforcesint.com/suppliers/defencevisionsys
tems.html
57. www.emccd.com/what_is_emccd
58. www.intevac.com
59. www.photonics.com/Article.aspx?AID=25144
60. www.flir.com/cvs/cores/view/?id=54460
61. www.photonis.com/en/ism/106l.html
62. www.jayceooi.com/2009/10/04/tftlcdvsoledwhichonet
ochoose/
63. www.printedelectronicsnow.com/articles/2012/06/emagins
amoleddisplaysenjoysuccessinmilitary
64. C.V. Driessche, F. Coursaget and F. Berthault, IRCMOS
fusion for night vision, 5th Int. Symp. on Optronics in De
fense and Security, Paris, 2012.
65. www.microoled.net/news/newsfrommicrooled
66. www.insighttechnology.com/productsinsight/fgs
67. airsoftinformations.blogspot.com/2010/04/envgenhanced
nightvisiongoggleanpsq.html
68. www.sidley.com/db30/cgibin/pubs/Statement_of_Facts.pdf
69. www.ulisir.com
70. www.globalsecurity.org/military/systems/ground/envg.htm
71. www.altaoptica.ru
72. www.obzerv.com
73. www.sensorsinc.com/LADAR.html
74. S. May, S. Fuchs, D. Droeschel, D. Holz, and A. Nuchter,
Robust 3Dmapping with timeofflight cameras, IEEE/
RSJ Int. Conf. on Intelligent Robots and Systems, St. Luis,
2009.
75. www.angenieux.com/zoomlenses/index.php?txt=38
76. J.L. Craig and H.L. Task, Development and evaluation of
the panoramic night vision goggle, Air Force Research Lab
oratory Human Effectiveness Directorate, Report DTIC no.
ADA400115, 1999.
77. Peter L. Marasco and H.L. Task, Optical characterization of
wide fieldofview night vision devices, Air Force Research
Laboratory Human Effectiveness Directorate, Report DTIC
no ADA430272, 1999.
78. Eric E. Geiselman and Jeffrey L. Craig, Panoramic Night vi
sion goggle update, Air Force Research Laboratory Human
Effectiveness Directorate, Report DTIC no ADA430243,
1998.
79. en.wikipedia.org/wiki/Night_vision_device
80. www.elbitsystemsus.com/sensorelectroopticssolutions/
land/soldier/helmetmounted/quadeye
81. www.insighttechnology.com/imageintensifiedsystems
products
82. www.nightvisionstore.com/Armasight%20SPARK%20Nig
ht%20Vision%20Monocular.htm
83. en.wikipedia.org/wiki/OLED
K. Chrzanowski
181

Review of Night Vision Technology

Uploaded by

Copyright:

Available Formats

Review of Night Vision Technology

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Review of Night Vision Technology

Uploaded by

Copyright:

Available Formats

What are image intensifier tubes and how do they work?

What are image intensifier tubes and how do they work?

What are the main components of a night vision device?

What are the main components of a night vision device?

OPTOELECTRONICS REVIEW 21(2), 153181

Review of night vision technology

of Optoelectronics, Military University of Technology, 2 Kaliskiego Str., 00908 Warsaw, Poland

Keywords: night vision technology, image intensifier tubes.

Humans achieve ability to see at night conditions by using

aviator night vision imaging system (a term used

Fig. 1. Block diagram of night vision monocular.

OptoElectron. Rev., 21, no. 2, 2013

Review of night vision technology

Over 99% of NVDs offered on world market can be

Fig. 3. Night vision monocular type Tarsius (Vectronix AG [5]).

OptoElectron. Rev., 21, no. 2, 2013

2013 SEP, Warsaw

Fig. 4. Night vision sight type Trident (ATN Corp. [2]).

Fig. 5. Night vision binoculars: (a) twochannel ATN Night Ra

The first two groups of NVDs (goggles, monoculars) are

In case of night vision monoculars the simplification

Review of night vision technology

3. Image intensifier tubes

1951 target of two tubes of different resolution at input

3.1. Generations of image intensifier tubes

Fig. 7. Magnified image of USAF 1951 resolution target generated

OptoElectron. Rev., 21, no. 2, 2013

2013 SEP, Warsaw

Fig. 8. Magnified image of USAF 195,1 resolution target generated

popular is division onto generations. US military has dic

Fig. 9. Spectral sensitivity of typical photocathodes: S1(Gen 0),

more sensitive than S25 photocathodes used in tubes of

Review of night vision technology

dows are used in Gen 1+ tubes to minimize degradation of

Fig. 11. Principle of work of a microchannel plate.

OptoElectron. Rev., 21, no. 2, 2013

2013 SEP, Warsaw

ricated from a lead glass, treated in such a way as to opti

sitivity of the first S25 photocathodes was about 250

OptoElectron. Rev., 21, no. 2, 2013

Review of night vision technology

Gen 3 tubes to be about two times better than for Gen 2

Fig. 14. Radiant sensitivity curves of Gen 2 tube photocathode

OptoElectron. Rev., 21, no. 2, 2013

2013 SEP, Warsaw

interest of U.S. military authorities and Night Vision and

using of autogating technique to power photocathode beca

Review of night vision technology

experiments with this technique were carried out for a de

3.2. Tube performance

Table 1. Basic parameters of IITs from different generations.

cascade inverter tube

inverter MCP tube

proximity focus MCP tube

proximity focus MCP, tube

proximity focus MCP, thin

proximity focus MCP,

OptoElectron. Rev., 21, no. 2, 2013

2013 SEP, Warsaw

tonis (formerly also DEP) introduced its own generation