Watch Your Perimeter
Video surveillance for perimeter security should pay close attention to light sensitivity and other IP camera features
- By Allen Chan
- Aug 01, 2016
When it comes to securing perimeters,
no two are exactly the
same. Perimeters can be vast
and remote like a national border,
or they can be small and
urban like the area surrounding a city bus terminal.
Intrusion concerns can be via land, air, water or all
three. Nearly any combination of factors is possible.
The one thing nearly all perimeters have in common
is that they are secured by some form of video
surveillance. This is generally done in combination
with other security measures, such as fences, guard
towers, sensors, radar, sonar or live patrols. Unlike
“open campus” environments like universities and
shopping malls, the goal of perimeter security is to
ensure that no one crosses a defined border without
being noticed.
This can be extremely challenging as security
breaches can be caused for any number of reasons, even
those that are completely innocent. For example, many
major airports are located adjacent to bodies of water
and the concern may be boats that unintentionally – or
intentionally – cross into airport areas. Because it may
not be feasible to install a fence, an innocent boater
may come ashore and into restricted areas.
In perimeter security, the purpose of a video system
is to see such occurrences and assess what response is
appropriate for the situation. When it comes to video
surveillance, there are a handful of technologies that
can help perimeter security meet its full potential.
LIGHT SENSITIVITY:
A NO. 1 PRIORITY FOR PERIMETER SECURITY
Light sensitivity is probably the single most important
specification an IP camera can have. Many perimeters
may not have the best lighting to cover the areas of
interest, which could pose a challenge when choosing
cameras. For example, a facility may be too close to
a populated area where light pollution at night is a
concern. Or, it simply may be impractical to install
lights due to the terrain or other factors. Therefore, a
camera must be able to produce good quality video in
whatever ambient lighting conditions exist.
At this point, many security professionals may be
wondering why light sensitivity is so important. After
all, the security industry has for many years offered
technologies that help cameras “see” better in
the dark, such as thermal imagers and infrared (IR)
light sources.
While thermal and IR imaging have their place,
both present challenges when it comes to video quality.
Over longer distances, objects may not be easy to
recognize due to resolution. They also essentially give
a “ghostlike” appearance to objects of interest, causing
figures to appear washed out and almost as if they
are glowing.
Using IR lighting requires that the object of interest
first be within the reach of the IR illuminator.
Long-range IR illuminators are available, but security
cameras are typically required to capture images of
objects that are both near and far. As objects get closer
to the camera and the IR illuminator, they become
overexposed because they are too close to the light source. This makes it difficult to use IR illumination to cover large areas with a single camera. In addition,
IR illumination causes white or light objects to show
up as black, and dark or black objects to show up as
white, making color identification difficult.
Thermal imaging does not have the distance restrictions
of IR, however, the color limitations still exist.
Warm objects appear white and cool objects appear
black, or vice versa. This makes it nearly impossible
to describe an object or person that may be breaching
a perimeter. In addition, even higher-end thermal
imaging uses VGA resolution, or standard definition,
imagers. This means the image is only 640x480, which
is extremely small. From long distances, it becomes
difficult to determine what an object is because it
is not possible to magnify the image. For instance,
it would be impossible to distinguish a deer from a
person once the distance surpasses the capabilities of
VGA resolution.
Thermal imaging works by detecting temperature,
and the scene is basically a gray scale of varying temperatures.
The heat signatures of objects with similar
temperatures appear the same, and if person or object
is in front of a structure or larger object of similar
temperature, the object in front might not be visible.
For example, if a person walks in front of a car with
a hot engine, the heat from the engine will mask the
person’s body heat and make it impossible for the
thermal imager to detect the person’s presence. This
tends to happen more in the summer months. A person
may be walking along a perimeter, but the heat
rising from the ground causes the surrounding area to
have a temperature similar to body heat. In this case,
the person can frequently “appear” and “disappear”
as the surrounding temperature changes.
We see things in color, describe things in color,
and so color plays a big factor in surveillance. Is the
car white, red, tan or blue? Is the person wearing a
green or black jacket? This kind of color information
is critical when a security breach is in progress. Color
information is also crucial forensic evidence for postincident
analysis or legal prosecution.
In contrast to IR and thermal imagers, IP cameras
that have excellent light sensitivity can continue to
create clear images with accurate color reproduction
as scenes darken and turn nearly pitch black. This produces better quality video without
the cost of extra equipment.
LIGHT SENSITIVITY AND
VIDEO RESOLUTION
In many perimeter security installations,
having the highest possible resolution
video is just as important as being
able to generate good images in the
dark. However, as camera resolution
goes up, light sensitivity goes down.
This is simply because when using
similar sized imagers, the sensor area to
gather light becomes smaller as resolution
increases.
This is particularly noticeable as users
move from full HD IP cameras to
megapixel or even 4K cameras. While
most HD cameras are perfectly capable
of generating decent color video
under only moonlit conditions, most
megapixel and 4K cameras have poorer
minimum illumination ratings than
their full HD counterparts. This means
they require more light to produce usable
images and will switch from color
mode to black-and-white mode before
an HD camera.
Better light sensitivity ratings for
megapixel and 4K cameras can be
achieved by using a full-frame 35mm
sensor, using a back-illuminated sensor
structure, using noise-reduction techniques,
or slowing down the camera’s
shutter speed to allow in more light.
When resolution is a key concern,
any of these can vastly improve the
camera’s overall light sensitivity characteristics
while still maintaining megapixel
and 4K resolution. In the darkest
conditions, 35mm sensors or back-illuminated
sensors will work much better
than cameras that try to increase gain
by using noise-reduction techniques or
slower shutter speeds. Unless the scene
is static, slower shutter speeds will increase
the blurriness of moving objects.
There are no standard specifications
for camera sensitivity in the security
industry, so even cameras that claim
to have the same ratings can perform
very differently. Therefore, it’s critical
for security professionals to do their
homework by asking the right questions
regarding how the minimum illumination
sensitivity figures are specified.
Better yet, onsite testing is the best
way to identify the megapixel and 4K
cameras with the best light sensitivity
performance, color reproduction, and
noise reduction.
LIGHT SENSITIVITY’S
IMPACT ON ANALYTICS
Having cameras with better light sensitivity
is also beneficial for any perimeter
security application that is running
analytics. When a camera’s images become
too dark, blurry or grainy, then
the accuracy of analytics software is
negatively impacted resulting in an increase
of false positives, or false alarms.
For example, high image quality is
needed in order for the analytics software
to determine if an object crawling
over a fence is a person or an animal,
or if a face matches one from the
watch list. If the object in question is
shadowed, or appears ghost-like from
thermal or IR imaging, then the analytics
software cannot work to its full
potential.
DYNAMIC RANGE AND LIGHT RESPONSE
Wide dynamic range (WDR) relates to light sensitivity,
but it’s not just for low-light conditions. The
dynamic range of a camera is rated in decibels (dB)
and determines the capability of the camera to see
the darkest and brightest areas of a scene at the same
time. The higher the rating, the better the camera
should operate in any difficult lighting condition, light
or dark. This can include excessively bright light from
the sun shining directly into the camera or high-contrast
scenes like a parking lot where some cars are in
the sun and some are shaded by trees. Dynamic range
technology allows clear images to be seen in light and
dark areas simultaneously, rather than overexposing
or underexposing various areas.
As with light sensitivity, there are no standards for
how a camera’s dynamic range is derived. Some products
can say 90dB, and some will say 100dB, but when
tested side by side, the 90dB performs better than the
one with 100dB. So, any products being considered
should be tested side by side to determine which performs
the best.
Another aspect to consider is how fast a camera
responds to sudden or extreme light changes. The response
characteristics of each product can be vastly different,
and unfortunately, there is no specification at all
for light response. Some cameras can take 10 to 15 seconds
to adjust to a sudden and drastic lighting change,
like from bright headlights flashing into the lens on a
dark night. In perimeter security applications, even a
few seconds could make the difference between detecting
an intruder and missing the security breach entirely.
A NOTE ABOUT VIEWING ANGLES AND ZOOM
Viewing angles are another important consideration
in perimeter security. Depending on the situation,
views may need to be wide and shallow, or narrow and
deep, so lenses should be chosen accordingly.
Similar to camera selection, choosing the right
lens comes with its own set of considerations regarding
light sensitivity. The longer (or deeper) the view
needs to be, the more light the lens needs in order to
generate the same quality image as a wider angle lens.
Increasing the f-stop even one increment will increase
the lens’ depth of field, but it also doubles its light requirements.
Therefore, security professionals need to
figure out how to either generate additional light, or
compensate with a more light-sensitive camera.
With the growing prevalence of 4K cameras in security
installations, it’s also important to keep in mind
a few things about 4K cameras, their viewing angles,
and their ability to zoom in on distant objects. Given
the same viewing distance and parameters, a 4K camera
produces video that is four times better than full
HD, and 27 times better than VGA. It can also allow
users to expand their wide area surveillance yet still
capture, magnify and examine the smallest parts of
a scene. But like other fixed IP cameras, a 4K camera
can be configured to see either wide angles or far into
the distance. It cannot do both at the same time. This
is where using a PTZ camera in conjunction with a
4K camera can be beneficial.
PERIMETER SECURITY FOR
TODAY AND TOMORROW
Without a doubt, securing perimeters is one of the
toughest jobs in the security industry. Each one is different
and presents a unique set of challenges. When
purchasing and deploying surveillance cameras, it is often
best to find the ones that can do the job, rather than
purely looking at costs. Just as with our daily life, using
the right tool for the job will make
things easier and faster, and it will be
more cost effective in the long run.
This article originally appeared in the August 2016 issue of Security Today.